This page contains a list of Master projects that are currently available in Life Sciences (SV) or in STI-IBI laboratories. These projects can be searched by keywords. If you are interested or would like to obtain further information from the Lab, please use the contact email listed.
Note that if you are interested in an interdisciplinary project (one where you will be co-supervised by two labs from different EPFL Schools), you can type ‘interdisciplinary’ in the search box.
| Laboratory | Title | Category |
| Lemaitre | Drosophila Immunity The Drosophila antimicrobial response at the time of the Cas9/CRISPR gene targeting revolution The application of Drosophila genetics has generated insights into insect immunity and uncovered general principles of animal host defense. These studies have shown that Drosophila has multiple defense “modules” that can be deployed in a coordinated response against distinct pathogens. Today, Drosophila can be considered as having one of the best-characterized host defense systems among the metazoan. The Cas9/CRISPR revolution offers new opportunities to revisit in a systematic manner Drosophila immunity. At the interface between large-scale genomic studies that lack resolution and individual gene analysis that lack breadth, our laboratory has undertaken a meso-scale ‘skilled’ analysis of immune modules, notably by addressing the individual and overlapping function of large immune gene family. The aim of the master project is to characterize the function of Drosophila immune modules (ex. antimicrobial peptides, phagocytosis,….) using powerful genetic approaches. Methods: Drosophila genetics, molecular biology, genomic, histology, microbiology, bioinformatic. Keywords: immunology, genetics, drosophila, antimicrobials Supervisor: Bruno Lemaitre Contact: [email protected] Required: A background in biology Availability: Open |
Infectious diseases Wet |
| Lemaitre | Insect endosymbiosis The foreign within: Drosophila-Spiroplasma interaction as a model of insect endosymbiosis Virtually every species of insect harbors facultative bacterial endosymbiotic bacterium (endosymbiont) that are transmitted from females to their offspring. Many manipulate host reproduction in order to spread within host populations. Others increase the fitness of their hosts by protecting their hosts against parasites. In spite of growing interest in endosymbionts, very little is known about the molecular mechanisms underlying endosymbiont-insect interactions. To fill this gap, we are dissecting the interaction between Drosophila and its native endosymbiont Spiroplasma poulsonii. The master project will use a broad range of approaches (molecular genetics, histology, microbiology, genomics….) to dissect the molecular mechanisms underlying key features of the symbiosis, including vertical transmission, regulation of symbiont growth, and symbiont-mediated protection against parasites. We believe that the fundamental knowledge generated on the Drosophila-Spiroplasma interaction will serve as a paradigm for other endosymbiont-insect interactions. Keywords: symbiosis, drosophila, bacteria, genetics Supervisor: Bruno Lemaitre Contact: [email protected] Required: a background in biology Availability: Open |
Infectious diseases Wet |
| Oricchio | Cancer Genomics data portal development The goal of this project is to organize and link cancer genomics data obtained from different sources and in different formats within a browsable data portal accessible to people in the lab. Keywords: Big genomic data, data portal, data integration Supervisor: Elisa Oricchio Co-supervisor: Giovanni Ciriello, UNIL Contact: [email protected] Required: python language required, knowledge of html/php preferred Availability: Open |
Computational Biology Interdisciplinary |
| Duboule | Hox gene regulation in cultured embryoids We have recently shown the implementation of collinear Hox genes expression in “gastruloids”, a mouse embryonic stem (mES) cells-based organoid that mimic early embryonic spatial and temporal genes expression [1]. The aim of this project is to generate stable deletions and knockin in mES cells in order to study the underlying mechanisms of Hox genes expression in this gastruloids. Keywords: Gene regulation, chromatin, organoids, CRISPR/cas9 Supervisor: Denis Duboule Contact: [email protected] Required: Some experience in cell culture and molecular biology is a plus Availability: Open |
Molecular biology Dry and wet |
| D’Angelo | Role of the Sphingolipid processing machinery in neural differentiation Starting from the 1970s, a number of studies have reported that the plasma membrane composition in terms of sphingolipids is subjected to remodelling during neural development. These changes are accompanied by a reprogramming in the expression of genes encoding the sphingolipid synthetic enzymes. Our preliminary data indicate that post-translational regulation of the sphingolipid synthetic machinery synergizes with transcriptional programs to assist GSL remodelling. Specifically we find that the localization and stability of key enzymes involved in the metabolic rewiring are inversely affected in the course of neural differentiation. Here we want to study the molecular mechanisms accounting for this post-translational regulation. Keywords: neural tube organoids, neural differentiation, lipid metabolism Supervisor: Giovanni D’Angelo Contact: [email protected] Required: – Availability: Open |
Developmental biology Wet |
| D’Angelo | Single-Cell in situ Lipidomics Lipids are fundamental constituents of all living beings. They participate in energy metabolism, account for the assembly of biological membranes, act as signalling molecules, and interact with proteins to influence their function and intracellular distribution. Eukaryotic cells produce thousands of different lipids each endowed with peculiar structural features and contributing to specific biological functions. Cellular lipidomes vary among cell types and are reprogrammed in differentiation events. Recent contributions including from our group have shown that lipid composition is subjected to remarkable cell-to-cell variation in syngeneic, homogeneous cell populations suggesting that cell-to-cell lipid heterogeneity contributes to the emergence of multicellular patterns. Lipid biologists have so far addressed lipidomes in bulk cell extracts or selected lipids at the single-cell level. Thus, how lipidomes vary form one cell to another and which cell-to-cell lipid variations have biological meaning remains to be defined. Here, we will develop an integrated pipeline coupling high-resolution mass spectrometry imaging, single-cell multi-omics and lipid probes to attain Single-Cell in situ Lipidomics analysis of cell populations. We will use this approach to interrogate the role of single-cell lipidome variations in the medically relevant case of dermal fibroblast heterogeneity.. Keywords: single cell omics, cell identity, senescence, cancer Supervisor: Giovanni D’Angelo Contact: [email protected] Required: – Availability: Open |
Cell biology Wet |
| Petersen | Brain wide single cell anatomy The goal of this project is to digitally reconstruct the axons and dendrites of cortical neurons in the mouse brain. Single neurons labelled with fluorescence will be imaged across the entire mouse brain at high resolution. The axons and dendrites will be traced through semi-automated procedures and quantified in the context of a standard mouse brain atlas. Keywords: mouse neocortex, axon, dendrite, 3D imaging Supervisor: Carl Petersen Contact: [email protected] Required: – Availability: Open |
Neuroscience Dry and wet |
| Petersen | Brain wide maps of inhibitory neurons The goal of this project is to comprehensively map the locations and numbers of genetically-defined types of inhibitory neurons in the mouse neocortex. Genetically-engineered mice will be used to label specific types of GABAergic neurons with fluorescent proteins, and then the entire mouse brain will be imaged at high resolution. The student will develop image analysis methods to locate each neuron within a standard mouse brain atlas. Keywords: Mouse neocortex, imaging, image analysis Supervisor: Carl Petersen Contact: [email protected] Required: Image processing Availability: Open |
Neuroscience Dry and wet |
| Petersen | Data analysis: electrophysiological recordings in behaving mice The goal is to analyse multi-site extracellular recordings of neuronal activity in mice performing a goal-directed behavior learned through reward-based training. The neurophysiological data will be correlated with behavior quantified from high-speed videography. Specifically, we will investigate how neuronal circuits in the mouse brain learn to transform whisker sensory information into goal-directed licking motor output through reward-based learning. Keywords: sensory processing, sensorimotor transformation, motor control, reward-based learning Supervisor: Carl Petersen Contact: [email protected] Required: Coding (Matlab or Python) Availability: Open |
Neuroscience Dry |
| Petersen | Data analysis: functional imaging in behaving mice The goal of this project is to analyse functional calcium imaging data in behaving mice. We will correlate the dynamic calcium signals with behavior quantified from high-speed video filming. The aim is to investigate how neuronal circuits in the mouse brain learn to transform whisker sensory information into goal-directed licking motor output through reward-based learning. Keywords: sensory processing, sensorimotor transformation, motor control, reward-based learning Supervisor: Carl Petersen Contact: [email protected] Required: Image processing, Coding (Matlab or Python) Availability: Open |
Neuroscience Dry |
| Petersen | Two-photon calcium imaging in behaving mice The goal of this project is to image neuronal activity in behaving head-restrained mice. Mice will be trained through reward-based learning to carry out a goal-directed sensorimotor transformation. Two-photon microscopy will be used to image neurons expressing genetically-encoded calcium indicators with cellular resolution. We will correlate neuronal activity with sensory stimuli and behavior quantified from high-speed filming. Students will need to follow Module 1 of the animal experimentation course or equivalent. Minimum project duration 6 months. Keywords: cellular imaging, sensory processing, motor control, reward-based learning Supervisor: Carl Petersen Contact: [email protected] Required: – Availability: Open |
Neuroscience Dry and wet |
| D’Angelo | Molecular Basis of COL4A3BP/CERT1 syndrome The human Ceramide Transfer Protein CERT (encoded by the gene COL4A3BP presently referred to as CERT1) is a cytosolic lipid transfer protein responsible for the non-vesicular transport of ceramide from the endoplasmic reticulum (ER) to the Golgi apparatus during sphingolipid biosynthesis. CERT1 activity is mediated by several functional motifs including an N-terminal pleckstrin homology (PH) domain that binds to phosphoinositide phosphatidylinositol-4-phosphate in the trans-Golgi and a C-terminal steroidogenic acute regulatory protein-related lipid transfer (START) domain that serves as a binding domain for ceramide. Genome-wide studies have suggested a putative association between CERT1 and intellectual disability (ID), still validation in a large cohort and dissection of the molecular etiology of the disease are lacking. We searched for human patients with CERT1 mutations and identified 19 individuals with de novo missense variants who suffer an infantile-onset developmental syndrome with severe ID, seizure and autism spectrum disorder. Here we want to address the molecular bases of the disease by studying the effects of patient mutations on CERT lipid transfer activity and on overall lipid metabolism. Keywords: Inborn errors of lipid metabolism, lipid transfer proteins, Membrane contact sites Supervisor: Giovanni D’Angelo Contact: [email protected] Required: – Availability: Open |
Cell biology Wet |
| Sakar | Engineering active biomaterials to study mechanical regulation of collective cell behavior Morphogenetic movements are generally believed to be guided by mechanical forces along with morphogen gradients. Likewise, emerging data show that cells respond to various mechanical signals including ECM stiffening due to deposition or remodeling of collagen fibers and compressive stress generated by confined growth. However, the physical mechanism causing such multicellular movements remains unknown. MICROBS Laboratory has been developing small scale soft actuated biomaterials that can transduce electromagnetic energy into mechanical work. We have recently introduced optomechanical microactuators that can apply physiologically relevant forces within a three-dimensional workspace. The objective of this project is the integration of these microactuators within biological models, by embedding into ECM with cells that form higher order structures such as spheroids and organoids. The student is going to work on the surface chemistry of the actuators to optimize the transmission of forces. A time-lapse fluorescence imaging methodology will be established that involves i) finding the excitation parameters (i.e. duration, frequency, amplitude of actuator contraction) that leads to a physiological (or pathological) multi-cellular response ii) development of techniques for high-throughput, multi-agent actuation schemes for the generation of arbitrary stress profiles, and finally iii) investigation of techniques for the mapping of stress throughout the tissue. Keywords: Bioengineering, mechnaotransduction, cell biology Supervisor: Selman Sakar Contact: [email protected] Required: Wet-lab experience Availability: Open |
Bioengineering Wet |
| Sakar | Engineering control over 3D morphogenesis During embryonic development flat, polarized epithelia sheets morph into complex three-dimensional structures and eventually form specialized compartments and organs. Initial bending and invagination events in epithelia are triggered by cellular contractile forces that lead to local cell shape changes and rearrangements. In principle, one can rationally shape biological tissues by controlling the location and magnitude of these mechanical forces. In this project you will work with engineered epithelial tissues and epithelial-mesenchymal multilayered constructs. First part of the project involves characterization of tissue composition and morphology using wide-field, confocal and light-sheet microscopy. In the second part, you will perturb cell mechanics by applying state-of-the-art manipulation technology and investigate local tissue architecture. Gained knowledge will be used to design and sculpt engineered 3D bodies. An in-house computational model will aid the exploration of the design space. Keywords: tissue engineering, microscopy, microtechnology, optogenetics Supervisor: Selman Sakar Contact: [email protected] Required: mammalian cell culture Availability: Open |
Bioengineering Dry and wet |
| Constam | Functional analysis of Activin-binding proteins The TGFβ-related prohormone Activin-A is frequently upregulated in solid tumors and mediates tumor-induced muscle wasting as well as oncogenic or tumor-suppressive functions, depending on the context. The goal of this project is to validate binding of Activin-A to specific proteins that we found in an interactome screen for factors that preferentially bind mature Activin-A or precursor processing intermediates, respectively, and to analyze their functions in regulating the bioavailability and local signaling of Activin-A in the tumor microenvironment to promote tumor immune evasion, or endocrine Activin-A signals that mediate muscle wasting. Keywords: TGFβ signaling, tumor immune evasion, proteomics, protein trafficking, proteases Supervisor: Daniel Constam Contact: [email protected] Required: Molecular or cell biology, or (bio)chemistry Availability: Open |
Cancer biology Wet |
| Lashuel | Recapitulating De Novo α-synuclein Aggregations in Astrocytes Astrocytes are known to provide various essential complex functions including structural and trophic support, primary roles in synaptic transmission, and maintenance of ionic homeostasis that allow efficient information processing by neuronal circuits (Nedergaard et al., 2003; Sofroniew and Vinters, 2010). Although alpha-synuclein (α-syn) is expressed at very low levels in astrocytes, α-syn inclusions are found in astrocytes under pathological conditions in the postmortem Parkinson’s disease brain (Wakabayashi et al., 2000; Song et al., 2009). As of today, there are no cellular or animal models that truly recapitulates astrocytic α-syn pathology as seen in postmortem PD brain. In the presence of α-syn overexpression, abundant α-syn aggregations could be detected (Lee et al., 2010; Sacino et al, 2014; Sorrentino et al., 2018) but the biochemical and ultrastructural properties of these aggregates and their relation to α-syn astrocytic pathology remains poorly understood. Since α-syn is expressed by human astrocytes and its protein level is a key determinant of aggregation, pathology formation and predisposition to developing PD, we plan to investigate conditions that could potentially induce changes in α-syn expression and its protein levels in primary astrocytes, with a greater emphasis on natural mechanisms rather than α-syn overexpression. Keywords: Astrocyte, Parkinson’s disease, Synuclein, Inclusion Supervisor: Hilal Lashuel Contact: [email protected] Required: Ideally basic knowledge with protein expression, primary neuron or astrocyte cultures Availability: Open |
Neuroscience Wet |
| Lashuel | Expression patterns of alpha synuclein and its pathological forms and their role in the pathogenesis of Parkinson’s disease A master student’s project is available in the Laboratory of Molecular and Chemical Biology of Neurodegeneration at the Swiss Federal Institute of Technology in Lausanne (EPFL), Switzerland (www.epfl.ch). The selected candidates will work on a project at the Laboratory of Molecular and Chemical Biology of Neurodegeneration (http://lashuel-lab.epfl.ch/) aimed to evaluate the expression patterns of alpha synuclein and its pathological forms in different brain areas and their role in the pathogenesis of Parkinson’s disease. To achieve this goal, the applicant will perform cutting-edge techniques such as iDISCO, that permits whole-mount immunolabeling with volume imaging of large cleared samples ranging from perinatal mouse embryos to adult organs, such as brains or kidneys. The desired applicant must be highly motivated to work as members of an interdisciplinary group working at the interface of chemistry and biology in collaboration with world-renowned research groups at the frontiers of neuroscience and brain research. The qualified candidate will benefit from working with very dynamic and multidisciplinary groups in a highly collaborative and stimulating environment and access to state-of-the-art laboratories and core-facilities. For more information about our groups, please visit our website and review our recent publications at http://www.ncbi.nlm.nih.gov/pubmed/?term=Lashuel. Keywords: parkinson’s disease, idisco, alpha-synuclein Supervisor: Hila Ashuel Contact: [email protected] Required: – Availability: Open |
Neuroscience Dry and wet |
| Constam | Regulation of mRNA translation by mechanosensory primary cilia Normal development and function of the gut and other internal organs such as our lungs and the heart depend on asymmetric activation of the Nodal cascade. In most deuterostomes, this process is governed by a directional flow of extracellular fluid that stimulates specific ion channels in primary cilia to repress translation of the secreted Nodal antagonist Dand5 specifically on the prospective left side. However, the mechanism that uncouples the regulation of mRNA translation from mechanosensory cilia in patients with laterality defects and other ciliopathies is unknown. The goal of this project is to elucidate how primary cilia activate the RNA-binding protein Bicc1 to repress Dand5 mRNA translation. Keywords: mechanosensation, cilia, translational regulation, Supervisor: Daniel Constam Co-supervisor: Andy Oates (SV) Contact: [email protected] Required: Molecular and cell biology, and/or confocal microscopy Availability: Open |
Developmental biology Interdisciplinary |
| Constam | Analysis of a novel consensus RNA binding motif of Bicc1 A tandem repeat of three RNA-binding K homology domains in Bicc1 can bind specific target mRNAs that are implicated in polycystic kidney diseases (PKD) and other ciliopathies. However, until recently, a consensus RNA sequence mediating these interactions and its role in Bicc1-mediated translational repression have remained unknown. This project will validate the importance of novel 4-nucleotide consensus motifs that emerged from an unbiased screen for Bicc1-interacting sequences in mediating the binding of reporter RNAs to recombinant KH domains, and in enabling Bicc1-mediated mRNA localization and translational repression in cell-free assays and in cultured cell lines. Keywords: translational regulation, RNA binding, reporter assays Supervisor: Daniel Constam Contact: [email protected] Required: Molecular biology Availability: Open |
Molecular biology Wet |
| Constam | Regulation of mRNA translation by inducible liquid-liquid phase transitioning Translational repression of specific target mRNAs by Bicc1 critically depends on its Sterile Alpha Motif (SAM) that mediates head-to-tail self-association of structurally well-defined protein surfaces in helical polymers. The goal of this project is to test whether the helical organisation is necessary for polymers to translationally repress Bicc1-associated mRNAs (e.g. by enabling Bicc1 to control RNA folding), and whether it can be engineered to regulate polymerisation at will, or whether the essential contribution of the SAM domain to mRNA silencing can be mimicked by alternative means of inducible self-aggregation other than SAM-mediated liquid-liquid phase transitions. Keywords: protein engineering, cell biology Supervisor: Daniel Constam Contact: [email protected] Required: (bio)chemistry and/or live imaging Availability: Open |
Bioengineering Wet |
| Lashuel | Investigation of seeding properties of mutant HTT proteins Huntington’s Disease (HD) is a genetic and progressive neurodegenerative disorder characterized by motor, cognitive and psychiatric symptoms. Despite the fact that the gene responsible of HD is known, the underlying mechanisms leading to huntingtin (HTT) aggregation, link to neurodegeneration and death is still not clear. Different neurodegenerative disease causing proteins are known to have prion like property. Using seeding property of recombinant alpha-Synuclein proteins, our lab was able to generate a neuronal model of PD and study process of Lewy formation in details (Anne-Laure et al. Biorxv, 2019). The objective of the project is to investigate the seeding property of mutant HTT proteins and the generation of a neuronal model of HD. For this, we will use either Pre Formed HTT Fibrils (PFF) or purified HTT inclusions from primary neurons (native seeds) and add them into primary neuronal culture to assess their ability to be uptaken and to seed the aggregation of endogenous HTT in neurons. The aggregation will need to be characterized by Immunocytochemistry (ICC) and biochemistry depending on different conditions: 1) Use of different HTT protein fragments; 2) Use of different polyglutamines repeats within the Htt protein; 3) Use of HTT protein with and without the Nt17 domain; and 4) The influence of PTMs on HTT protein. Keywords: Huntington’s disease, Primary neurons, aggregates, seeding Supervisor: Hilal Lashuel Contact: [email protected] Required: Ideally basic knowledge with Biochemistry, imaging and cell culture. Availability: Open |
Neuroscience Wet |
| Sakar | Mechanics of tail elongation in zebrafish embryos The morphogenesis of vertebrate embryos is well-studied in terms of genetics and biochemical signaling. However, the role of mechanics is not well understood, primarily due to lack of tools and methods to apply forces at relevant spatiotemporal scales. To have a better understanding of the role of mechanics, we have recently developed an ex-vivo assay. Objective of this project is to perturb and probe samples with various micromanipulation systems. The systems will enable application of physiologically relevant stresses on the explants and quantify the results on somitogenesis and morphing. Optimizing the platforms, development of proper imaging tools for recording oscillatory gene signals, and post processing the data for quantification are the main elements of this project. Keywords: mechanobiolgy, somitogenesis, zebrafish, microengineering, robotics` Supervisor: Selman Sakar Co-supervisor: Andy Oates, SV Contact: [email protected] Required: basic knowledge in solid mechanics Availability: Open |
Developmental biology Interdisciplinary |
| Schneggenburger | Investigating the effect of GCaMP6 expression on plasticity in the neural circuits involved in fear learning Genetically encoded calcium indicators (GECIs) are powerful tools for monitoring intracellular calcium concentration and thus, indirectly, neuronal spiking activity. The family of GCaMP6 GECIs has become instrumental in neuroscience, particularly in in-vivo recordings. However, it is often disregarded that GECIs, like other Ca2+-indicators, are calcium buffers, usually of high affinity, which in typical experiment are strongly overexpressed in every cellular compartment. The presence of a strong buffer is likely to affect naturally occurring calcium-dependent plasticity mechanisms. Our laboratory is setting up an in-vivo calcium imaging approach using a microendoscope to study population dynamics in the insula cortex and amygdala during fear learning. It is therefore important to know if/how GCaMP6 expression influences the function of studied neurons in memory-related tasks. The project will first focus on estimating the added calcium buffering capacity in neurons of lateral amygdala (LA) expressing GCaMP6m using fluorescent calcium imaging and patch-clamp in brain slices. Second, the long-term plasticity of synaptic inputs from the insula cortex will be investigated in the LA neurons expressing GCaMP6m using patch-clamp electrophysiology. These experiments will allow the MA student to investigate quantitative aspects of GCaMP6m overexpression and to study how this affects neuronal plasticity. Keywords: synaptic plasticity, calcium imaging, patch-clamp electrophysiology, brain slices Supervisor: Ralf Schneggenburger Contact: [email protected] Required: completion of neuroscience/neurobiology course Availability: Open |
Neuroscience Wet |
| Schneggenburger | Development of an automated pipeline for brain-wide counting and reference atlas registration of fluorescently labelled neurons In our laboratory, we study the neural networks involved in associative learning of threat in auditory fear conditioning. One of approaches studies the distribution of neurons in the brain that undergo increase in activity upon specific sensory experience. This is achieved using a fluorescent reporter mouse line combined with timed expression of Cre-recombinase downstream of an immediate early response gene cFos. A typical outcome is a set of hundreds of images of histological sections of the mouse brain, in which the distribution of “activity-labeled” neurons needs to be quantified. This is an extremely workload-demanding task in case of manual analysis. There have been multiple recent attempts to automate the detection of single cells and registration of the images onto a reference atlas using advanced image analysis and deep learning. However, there is no single comprehensive pipeline accepting an image stack at its input and producing the cell density table at its output with minimal human input. The project is devoted to a design of such a workflow based on further development of interfaces between some of the published routines. The resulting toolbox will be very instrumental for our lab in particular, and for neuroscience community in general. Keywords: image analysis, object detection, image registration, automation Supervisor: Ralf Schneggenburger Contact: [email protected] Required: advanced programming in MATLAB, basic knowledge of image analysis Availability: Open |
Computational Biology Dry |
| Hummel | Investigation of resting-state fMRI network changes following noninvasive brain stimulation during sleep in elderly and stroke patients Objective: Sleep is essential for learning new skills at any age and for relearning functional capacities (such as motor skills) after a brain lesion (stroke). In the context of an ongoing project aiming at investigating the effects of non-invasive brain stimulation during sleep to enhance learning, the master student will analyse a resting-state fMRI dataset of healthy older / stroke patients. The project will consist in acquisition of data and analyses of structural and functional MRI data. Furthermore, the analysis will include electric field modelling of the applied brain stimulation to gain insight on the stimulation effect and topography. Keywords: Resting-state fMRI, Noninvasive brain stimulation, Electric field simulation, Multimodal study Supervisor: Friedhelm Hummel Contact: [email protected] Required: – Availability: Open |
Neuroscience Dry |
| Hummel | fMRI default mode network changes after stroke. The brain is never at rest. Whatever action we may undertake, the brain is processing sensory information to act on the environment. Particularly, it was discovered in recent years that even by being at rest specific areas in the brain are active within a default mode network (DMN) comprising of the posterior cingulate cortex (PCC) or the ventral Anterior Cingulate Cortex (vACC), thought to play a role in different brain functions, as self-reference, social evaluation or episodic memory. However, the relationship of the DMN with different pathological clinical states after stroke is not well defined. After a stroke, lesions may cause damages at different cognitive levels: motor, language, attention etc. In a longitudinal study after stroke, multimodal data are collected from patients suffering from upper limb deficit after stroke. During 4 time points from 1 week to 1 year patients well be characterized with MRI scans for structural and functional MRI analyses. The goal of the study is to evaluate the correlation of the DMN activity with stroke outcome and functional reorganization informed by the structural connectome of each individual patient. Keywords: stroke, fMRI, brain imaging, default mode network Supervisor: Friedhelm Hummel Contact: [email protected] Required: – Availability: Open |
Neuroscience Dry |
| Hummel | Longitudinal stroke evaluation and graph analysis at rest from EEG Stroke injury leads to disability up to 75% of stroke survivors. The understanding of the processes underlying recovery of stroke and the plastic correlates in the brain are not sufficiently understood. Electroencephalography (EEG) provides excellent information about brain connectivity with a high temporal resolution and will add to the understanding of the brain correlates of recovery from stroke. Advanced analytical techniques based on graph analysis will be used to investigate brain activity at rest. To this end, in a longitudinal study multimodal data are collected from stroke patients suffering from upper limb deficit. Patients will be evaluated from the acute (1 week) to the chronic stage (1 year) at 4 time points, brain activity is recorded by means of EEG. The goal of the study is to characterize the changes of the functional connectome during the process of recovery Keywords: stroke, EEG, graph analysis, longitudinal study Supervisor: Friedhelm Hummel Contact: [email protected] Required: – Availability: Open |
Neuroscience Dry |
| McCabe | Robogenetics: Genome engineering for robotics The goal of this interdisciplinary joint lab project is to automate Drosophila genetics and mutant selection through a combination of robotics with new genetic tools designed for machine rather than human manipulation. Studies of the fruit fly Drosophila melanogaster have made foundational contributions to the understanding of development, neuroscience and human disease. Existing proof-of-concept robotic systems can anesthetize, transfer and manipulate individual Drosophila. The student will build and adapt such a system to develop a new generation of ‘smart’ robots with sensors, which together with new genetic tools, will allow for robotic selection of animals of the desired genotype or mutation. This combination of novel genetic tools designed specifically for machine utility with robotics should facilitate an exponential increase in the throughput of genetic manipulation of this important model organism. Keywords: genetics, robotics, machine vision, genome engineering Supervisor: Brian McCabe Co-supervisor: Pavan Ramdya (IBI, SV) Contact: [email protected] Required: – Availability: Open |
Neuroscience Interdisciplinary |
| McCabe | Non-invasive assays of neurodegeneration in models of human disease Loss of synaptic connections and disruption of neuronal circuits is thought to be an early consequence of neurodegenerative disorders such as Alzheimer’s disease. However, assaying these early aspects of neurodegeneration is difficult and currently cannot be achieved in vivo. In this project, the student will use genome engineering methods to construct new genetic tools for the model organism Drosophila melanogaster designed to measure neural circuit integrity non-invasively from outside the animal. They will then deploy these tools in a humanised model of Alzheimer’s disease to assay the progressive disruption of neuronal circuits as neurodegeneration advances. Keywords: genome engineering, neurodegeneration, disease, Drosophila Supervisor: Brian McCabe Contact: [email protected] Required: – Availability: Open |
Neuroscience Wet |
| La Manno | Extensions of the RNA velocity algorithm Background An ambitious goal of single-cell analyses is describing dynamical biological processes and shedding light on gene regulation mechanisms. However, because of the destructive nature of single-cell measurements, they can only provide a static snapshot of the cell at a given point of embryonic development. To overcome this fundamental limit, I recently developed a novel method, named “RNA velocity” as it estimates the first derivative of gene expression for each gene in a cell (RNA velocity of single cells, Nature 2018). Measuring the abundance of both unspliced and spliced RNA in the same cell, we can estimate the rate of change of gene expression and predict the future expression levels of a single cell. Activities This project will start by introducing a series of improvements to the current RNA velocity algorithm and software and it will proceed towards an extension of the model to a more general framework. (1) The student will implement in our software velocyto with the dynamical model proposed recently by Fabian Theis lab. (2) The student will analyze situations where the original assumptions of the method are not met and remediate fitting an alternative model (3) The student will use a latent variable model to decompose the velocity vector in interpretable components (e.g. cell cycle, maturation, and response to signals). Keywords: machine learning, developmental biology, single-cell RNAseq Supervisor: Gioele La Manno Contact: [email protected] Required: Experience in numerical python programming (numpy, scipy, matplotlib), at least a course related to either multivariate statistics or machine learning Availability: Open |
Computational Biology Dry |
| La Manno | Implementation and extension of multimodal data integration algorithms Background The scientific community is generating a great volume of datasets that record many features from single cells. These datasets survey the same population but each dataset is measuring different features: transcript, chromatin accessibility, DNA-methylation, surface protein concentration. These measurements are disjoint and do not come from the same cell. However, since the datasets are collected from the same organ/tissue/process it should be possible to align this data to obtain a more complete description of the molecular state of each cell. This kind of approach has proved to be possible through a series of multivariate statistics/machine learning procedures (Stuart et al. 2019). Activities The project is divided into two parts an implementation and a method extension part. In the first part, the student will individuate in the literature methods for multimodal data integration and batch correction and reimplement the best of them in python. In the second part, the student will identify common and different procedures in those methods and attempt to combine different them so as to propose a new improved version. Finally, the different methods including the improved one will be benchmarked using different reference datasets. Keywords: single-cell transcriptomics, bioinformatics, data analysis Supervisor: Gioele La Manno Contact: [email protected] Required: Knowledge of python (good) and R (basic) languages. Some experience with multivariate data analysis. Having implemented a machine learning method. Availability: Open |
Computational Biology Dry |
| La Manno | Data-aware models for the design of maximally informative high throughput experiments Background Single-cell RNA-sequencing (scRNAseq) yields high-quality transcriptomic data from a single cell suspension, but it needs to sacrifice information on cell localization. Conversely, multiplexed single-molecule FISH (osmFISH) has only medium-throughput but allows measurements in situ. Usually, after scRNAseq has been used to define and discover cell types, one would like to map back this knowledge to the tissue using osmFISH. In doing so, it is important to be able to optimally select a gene set that maximizes the information obtainable experimentally. For example, choosing 30 highly specific markers to detect 30 cell types might not be the best option, because, without redundancy, the failure of a probe could mean completely missing a population. We want to design a method that automatically chooses a subset of features that can be highly informative while robust to this kind of experimental uncertainty. Activities The student will (1) Build a model that predicts dense low-throughput measurements from sparse high-throughput data using machine learning and statistical modeling. (2) Design a strategy that optimally selects a set of features to measure by a low-throughput method. (3) Evaluate performances of the method considering uncertainties (4) Incorporate experimenter knowledge in the selection in the form of statistical priors. (5) integrating the procedure in a command-line tool. Keywords: machine learning, statistical modeling, algorithms Supervisor: Gioele La Manno Contact: [email protected] Required: python programming, familiarity with statistical modeling and multivariate regression Availability: Open |
Computational Biology Dry |
| La Manno | Probing the influence of organizers on the transcriptome of neural stem cells Background It is becoming increasingly evident that radial glia-like cells, the stem cells of the nervous system, are not a homogeneous population. In particular, using single-cell RNAseq we have identified different molecularly distinct states corresponding to spatiotemporal patterning of the brain. We would like to understand both the causes and the functional implications of the pattern observed. We are interested in investigating how the microenvironment and specific organizers influence or determine this pattern. We expect that paracrine signals are important in inducing those non-autonomous expression changes. Activities The student will culture primary cells obtained from embryonic mice brains at different ages and regions, and evaluate the composition of the cell types generated in vitro. Cocultures will be set up with “majority” cell population from a region, and a “minority” cell population derived from another region. The idea is that the bigger fraction of cells will provide a dominating amount of signaling molecules. Single-cell RNA sequencing will be performed on these cells, allowing a thorough comparison at the transcriptome level. Keywords: single-cell transcriptomics, neuroscience, stem cells Supervisor: Gioele La Manno Contact: [email protected] Required: Some experience with culturing cells, basic python programming Availability: Open |
Developmental biology Dry and wet |
| van der Goot | Microtubules and organelle architecture We are interested in understanding how the complex architecture of the endoplasmic reticulum is established and maintained. We recently identified a novel protein, through a proteomics approach, that connects the ER to microtubules. The project will be to investigate how this protein contributes to controlling ER shape and ER dynamics. It will involve using a variety of approaches, in particular super resolution microscopy, molecular biology and biochemistry. Keywords: cell biology, molecular biology, bio engineering Supervisor: Gisou van der Goot Co-supervisor: [email protected] Contact: [email protected] Required: – Availability: Open |
Cell biology Interdisciplinary |
| van der Goot | Understanding the cellular dynamics of endoplasmic reticulum (ER) shaping proteins CLIMP63 is a crucial regulator of ER architecture. It controls the dynamics and the compartmentalisation of the Endoplasmic reticulum. Using high resolution microscopy we intend to follow the dynamic of various CLIMP63 mutants to understand how the modification of CLIMP63 by lipids can reversibly tune ER morphology and possibly interfere with ER interactions with other cellular organelles. Keywords: Cell and molecular biology, Bioengineering, microscopy Supervisor: Gisou van der Goot Co-supervisor: [email protected] Contact: [email protected] Required: – Availability: Open |
Cell biology Interdisciplinary |
| van der Goot | Structural determination of a key ER shaping protein The ER forms a dynamic network of morphologically distinct compartments, including the nuclear envelop, the rough ER and smooth ER tubes. The morphology of the rough ER is to a large extent controlled by the transmembrane protein CLIMP63. To understand how CLIMP63 shapes membranes, we will study its 3D structure and its quaternary assembly. We will investigate both the lumen domain and the full length protein, which will be reconstituted into artificial lipid bilayers. The project will involve various techniques: protein purification, electron microscopy, biochemistry Keywords: Cell and Molecular biology, Bioengineering, Biophysics Supervisor: Gisou van der Goot Co-supervisor: [email protected] Contact: [email protected] Required: – Availability: Open |
Cell biology Interdisciplinary |
| Persat | The mechanics of the gut microbiome The main goal of this project is to determine if and what gene expression changes after bacterial cells contact the surface that resemble the ones of human cells. In this project, we will consider a variety of bacterial species and identify their response to surface contact. We will both subject bacterial populations of synthetic surfaces such as hydrogels but also to biological tissue-like systems generated from organoids. You will then extract RNA from these cells and measure their gene expression using state-of-the-art RNA-seq technology. Finally, you will use bioinformatic tools and statistics to determine how different species might adapt to the surface attachment. Overall, you will work with a wide variety of tools and technologies, including next-generation sequencing, bioinformatics, microfluidics and tissue-engineering. Keywords: gut microbiome, organoids, mucus, mechanics Supervisor: Alexandre Persat Contact: [email protected] Required: – Availability: Open |
Bioengineering Wet |
| Persat | Characterizing mechanosensation by cryo electron microscopy Living systems sense and respond to mechanical cues present in their environments. For example, mechanics influence stem cell differentiation, development, motility and cancer progression. In bacteria, mechanics play an important role in regulating pathogenicity. Our lab focuses on understanding how bacteria sense and respond to forces. We have identified bacterial structures that are sensitive to forces, and now must capture the structural changes in these structures. In this project, you will design a new technique to apply forces on bacteria directly on electron microscopy grids, ultimately activating mechanosensors and allowing us to decipher their structural dynamics. This is an exciting project that uses methods from engineering (microfluidics, materials, afm), physics (light and electron microscopy) and biology (microbiology, mechanobiology), but no prior knowledge is required. Keywords: Supervisor: Alexandre Persat Contact: [email protected] Required: – Availability: Open |
Bioengineering Wet |
| Biorobotics Laboratory | Neuromechanical simulations of animal and human locomotion The biorobotics laboratory has several semester and master projects in neuromechanical simulations of animals and humans. The project descriptions can be found here: https://biorob2.epfl.ch/pages/projects/ Keywords: locomotion, numerical models, biomechanics Supervisor: Auke Ijspeert Contact: [email protected] Required: Programming in Python and C Availability: Open |
Neuroscience Dry |
| Suter | Quantitative analysis of protein synthesis and degradation in Zebrafish by fluorescence microscopy How protein levels are regulated by the interplay of protein synthesis and degradation is still poorly understood. We recently developed a fluorescent timer-based approach allowing to simultaneously monitor protein synthesis and degradation in individual, living cells. This project will be based on the expression of this fluorescent timer in zebrafish to monitor protein synthesis and degradation by quantitative fluorescence imaging. It will involve various techniques such as DNA cloning, zebrafish transgenesis, conventional/light sheet fluorescence microscopy and quantitative image analysis. Keywords: protein homeostasis, fluorescent timer, zebrafish, live imaging Supervisor: David Suter Co-supervisor: Andy Oates Contact: [email protected] Required: The student should be highly interested in microscopy image analysis and in gaining a quantitative understanding of gene expression. Availability: Open |
Cell biology Interdisciplinary |
| Herzog | Serial dependence in visual perception Our recent experience has a strong impact on how we perceive the present: the daylight appears brighter after leaving a dark room, a white surface appears greener after staring at a red image. In some cases, prior events may even deceive us to perceive the present as more similar to the past than it actually is, a phenomenon known as serial dependence. The objective of this project is to investigate the role of prior experience and temporal dependencies in human vision by combining behavioral psychophysics and computational modeling. The student will be involved in the collection and analysis of psychophysical data with the primary objective to develop a physiologically plausible computational model of serial dependence in human vision. Keywords: vision, computational modeling, psychophysics Supervisor: David Pascucci Co-supervisor: David Pascucci Contact: [email protected] Required: Good analytical and computational skills, ideal for students in: Computer Sciences, Life Sciences, Bioengeneering Availability: Open |
Neuroscience Interdisciplinary |
| Radenovic | 3D parameter free resolution estimation Recently we published a new method for image resolution estimation. The method is parameter free and exploits the phase information contained in the Fourier space of the image. Through the calculation of many partial phase correlation of the image with a filtered version, we are able to reliably extract the frequency support of any image, that is the highest frequency with significant contrast with respect to noise. So far we demonstrated the ability of the method to estimate the global, local and sectorial resolution of various cell samples and we would like to extend the method in the third dimension. This extension presents several problems that the student will have to understand and overcome, such as limited axial sampling, plane to plane coregistration, increased computational complexity, etc… The task of the student will be to implement the 3D version of the algorithm based on the current implementation (first in Matlab and then Java). The student will also have to investigate several options for processing optimization (code refactoring, GPU implementation, minimizing the number of correlations to be computed) due to the large amount of data to be processed inherent to volumetric imaging. Keywords: Resolution estimation of 3D imaging datasets Supervisor: Aleksandra Radenovic Contact: [email protected] Required: Knowledge in signal and image processing, Matlab, Java and basics of optical ima Availability: Open |
Bioengineering Dry and wet |
| Radenovic | Deep learning assisted segmentation and mapping of DNA molecules DNA analysis methods have evolved tremendously over the past decade. One of the goal of such techniques is to be able to recognize the species of origin. As an alternative to DNA sequencing (i.e. reading the whole DNA sequence), we have developed in our lab a way to map the DNA to its corresponding species while avoiding complicated PCR reactions and DNA sequencing. The method is based on sequence specific labelling of DNA and subsequent stretching on a glass surface. The stretched DNA is then imaged with a super-resolution microscope resulting in a sort of bar-code image (Figure). The intensity profile of each DNA molecules is extracted and matched against a database of species. 1,2 In order to study the entire microbiome, we need to analyse thousands of images, extract all the individual DNA molecules and match them to their sequences. This is too much for manual selection, a method is needed to automatically detect the DNA strands and extract their intensity profile. The task of the student will be to optimize a new approach to DNA segmentation based on machine learning and work on the automatisation of the full pipeline, from raw images to meaningful information. The second task of the student (Master project) will be to train another neural network to classify the segmented DNA. We will provide the student with experimental images, supervision and expertise to develop the algorithm (Python and Matlab). The student will be able to work in a highly interdisciplinary group with backgrounds ranging from polymer physics, image analysis, microscopy to molecular biology. Keywords: Image processing of super-resolved stretched DNA Supervisor: Aleksandra Radenovic Contact: [email protected] Required: Knowledge in signal and image processing, machine learning and basics of optical imaging, Availability: Open |
Bioengineering Dry and wet |
| Radenovic | 3D Super-Resolution Optical Fluctuation Imaging for Molecular Parameter Estimation Several techniques have been developed in order to overcome the diffraction limit in fluorescence microscopy. They rely on exploiting a priori knowledge about the quantum mechanical properties of fluorophores. For super-resolution optical fluctuation imaging (SOFI), higher order statistics (cumulants) of a time series of blinking fluorescence emitters are computed. SOFI analysis yields essentially background free images and is ideally suited for fast, 3D imaging using a multi-plane microscope. Balanced (b)SOFI deals with the nonlinear image contrast and enables up to fivefold improved spatial resolution in 2D. It allows extraction of molecular parameter maps such as the density and state lifetime of molecules. Molecular counting (or density estimation) based on SOFI is robust against imaging artifacts, which makes SOFI a unique tool in the super-resolution microscopy research field . We recently used bSOFI to investigate clustering of membrane proteins in T cells with an aim to understand their response In this project, you will explore the extension of bSOFI parameter estimation to 3D imaging. Based on an existing 2D simulation framework, different multi-plane configurations and fluorophore properties can be simulated (brightness/ signal-to-noise, blinking parameters, photo-bleaching) and evaluated using SOFI processing. You will perform super-resolution imaging of cells and/or DNA origami/protein calibration standards on our custom multi-plane microscopes. We aim to demonstrate multi-plane bSOFI parameter estimation for the first time. Depending on your skills and interests, the in silico or experimental aspects of the project can be emphasized. Keywords: Super-resolution imaging Supervisor: Aleksandra Radenovic Contact: [email protected] Required: Optics, microscopy, data analysis (Matlab), signal processing, cell culture, labelling of cells and preparation for super-resolution microscopy Availability: Open |
Bioengineering Dry and wet |
| Radenovic | Optical Projection Tomography to Elucidate Neurodegeneration Our research is set in the context of studying Alzheimer’s disease and observing Amyloid-beta plaques and immune cells using the 3D whole-tissue imaging modality known as OPT (see Nguyen, D. et al. 2017 Biomed Opt Express). The project’s aim is to apply OPT imaging to identifying amyloidosis levels and immune response in the brain and the gastro-intestinal tract of the 5xFAD mouse model. The task of the student will entail collaborating with microscopists and biologists to optimize the resolution optics of our existing in-house OPT setup while in parallel testing various antibody-mediated staining of tissue sections and whole tissues. All in vivo work will be performed by appropriately licensed staff members, and thus will not be part of this student project. The student will be able to work in a highly interdisciplinary group with backgrounds ranging from polymer physics, image analysis, microscopy to molecular biology. Keywords: study of neurodegeneration in a mouse model of Alzheimer’s disease using OPT Supervisor: Aleksandra Radenovic Contact: [email protected] Required: Basics of optical imaging, basic engineering techniques (e.g. soldering), general biology knowledge Availability: Open |
Bioengineering Dry and wet |
| Mathis | Anatomical connectivity of the neural circuits required for learning motor skills The lab studies the neural basis of motor learning by training mice to play skilled games. We have shown that cortex is important for this type of learning (Mathis et al 2017 Neuron). We now aim to study the underlying anatomical connections between cortex, brainstem, cerebellum, and the spinal cord. This project will leverage cutting-edge imaging of neural circuits (mesoSPIM) and anatomical tracing coupled with deep learning tools for efficient data analysis to elucidate the interconnectivity of important nodes for motor learning. Experience with stereotaxic surgery, viral injections, and tissue processing is highly desired. The deep learning-based tools can be learned. Minimum project duration of 6 months. Keywords: stereotaxic surgery, anatomy, deep learning, neuroscience Supervisor: Mackenzie Mathis Contact: [email protected] Required: Experience with mice, programming with Python Availability: Open |
Neuroscience Dry and wet |
| Mathis | Task-driven hierarchical deep neural networks model the proprioceptive pathway Proprioception is critical for sensing and controlling the body, yet no canonical hierarchical model of the system exists. Task-driven modeling has provided important insights into other sensory systems. However, unlike for vision and audition, databases of relevant proprioceptive stimuli are not readily available. For this project the goal is to design different tasks (and create training datasets) as well as train models of the proprioceptive pathway on various behavioral tasks. We will then analyze the networks’ units as well as emerging computations to gain insights into proprioception. We are specifically interested in understanding the implications of different tasks and network architectures. Keywords: task-driven modeling, deep learning, proprioception, motor control Supervisor: Alexander Mathis Contact: [email protected] Required: Python programming skills • Knowledge in TensorFlow/PyTorch would be great (not required) Availability: Open |
Neuroscience Dry |
| Mathis | Deep action recognition networks for animal behavior analysis We strive to develop tools for the analysis of animal behavior. Behavior is a complex reflexion of an animal’s goals, state and individuality. Thus, accurately measuring behavior is crucial for advancing basic neuroscience, as well as the study of various neural and psychiatric disorders. The goal of the master thesis project is to optimize deep neural network architectures to predict the behavioral state of animals in various experiments with limited amounts of data. For this project various datasets from our collaborators will be utilized. Keywords: Action recognition, deep learning, Behavior Supervisor: Alexander Mathis Contact: [email protected] Required: Python programming skills • Knowledge in TensorFlow/PyTorch would be great (not required) Availability: Open |
Neuroscience Dry |
| Jaksic | Simulating experimental evolution of cognition Experimental evolution is a long-term experiment that is highly dependent on the initial experimental design. Hence, optimizing the experimental design for most efficient selection outcome is imperative. Using population genetics theory and forward simulations we can predict the quantitative outcomes of different experimental designs. This allows us to choose the most fitting experimental design while also producing an empirical null expectation from the experiment for future hypothesis testing. Optimal evolution experiment design yields highest resolution for identifying adaptive loci in the shortest amount of time. This is especially important for complex phenotypes that are expected to be influenced by many genetic loci of small effect which obfuscate the signal of selection. In our lab, we are interested in the genetic basis of evolution of cognition, arguably one of the most complex traits we know of. In this project you will use genotype data from sequenced Drosophila melanogaster lines to create a virtual population and simulate its different evolution scenarios. The results will help us decide on the design and interpretation of the first evolution experiment for cognitive ability. Keywords: population genetics, experimental evolution, simulation Supervisor: Ana Jaksic Contact: [email protected] Required: R; Basic command line scripting; Data formatting using awk, grep or similar Availability: Open |
Computational Biology Dry |
| Jaksic | Genetic basis of the interplay between dopamine and octopamine Dopamine is a neurotransmitter important for regulating many behavioral traits, including locomotion. We previously screened for genome-wide genetic variation that underlies natural variation in locomotor ability in Drosophila caused by dopamine level perturbation. This uncovered a potential interplay between octopamine and dopamine in balancing external perturbations to these neurotransmitters. The aim of this project is to functionally validate the effects of octopamine signaling genes on locomotor variation caused by imbalance in dopamine levels. You will use Drosophila melanogaster GAL4>UAS transgenic system for targeted knock down of candidate gene expression in combination of pharmacological perturbation of neurotransmitter levels, and assay the effects on locomotor ability of flies. This outcome of this project will allow us to explore how the two neuronal signaling systems interact on a genetic level. Keywords: genetics, neurobiology, drosophila Supervisor: Ana Jaksic Contact: [email protected] Required: basic genetics, willingness to work with flies, willingness to learn Availability: Open |
Neuroscience Wet |
| Manley | The when and wheres of mitochondrial gene expression Mitochondria are the reason you breathe. They enable efficient burning of sugars and fat to produce ATP by cellular respiration. This in turn depends on the expression and maintenance of their own DNA (mtDNA). In this project you will aim to elucidate the distribution of mtDNA and how its replication and transcription are regulated. You will engineer cells to express transcription and replication markers, perform live and fixed cell confocal and super-resolution fluorescence microscopy, and finally analyse and evaluate the image-data. Ultimately, a better understanding of mtGene expression might provide the basis for to target many mitochondria-related diseases, such as f.i. Alzheimers. Keywords: mitochondrial DNA, fluorescence microscopy, human cells Supervisor: Suliana Manley Contact: [email protected] Required: enthusiasm, independence, self-discipline, capable of having fun Availability: Open |
Molecular biology Dry and wet |
| Sandi | Analytical approaches to data from human virtual reality (VR) and neuro-physiology studies The combination of Virtual Reality (VR), motion tracking, autonomic response recording and EEG provides a versatile and information enriched way to study human behavior and neurophysiology in a laboratory setting. The analysis of the resulting multivariate datasets is a challenging task requiring a combined knowledge from different areas like signal processing, machine learning, statistics and data visualization. Keywords: Supervisor: Carmen Sandi Co-supervisor: joã[email protected] Contact: [email protected] Required: Both pre-processing and analysis of the output multivariate datasets comprised of full body motion, eye tracking, pupil dilation, respiration, heart rate, skin conductance, EMG, and EEG. And programming skills (Matlab and/or R and/or Python). Availability: Open |
Neuroscience Interdisciplinary |
| Sandi | Development of an application for multimodal autonomic biofeedback The student will contribute to the development and piloting of a biofeedback application that integrates physiological signals (heart rate, heart rate variability, skin conductance and breathing rate). The aim is to study the effectiveness of our protocol to reduce anxious behavior or stress responses. Feedback can be delivered on a computer screen or later, via an immersive environment through a head mounted display, depending on the student’s progress. Keywords: Supervisor: Carmen Sandi Co-supervisor: [email protected] Contact: [email protected] Required: Basic knowledge in the aforementioned techniques and programming skills (Matlab and/or R and/or Python) Availability: Open |
Neuroscience Interdisciplinary |
| Sandi | Effects of stress and anxiety on motivated behaviour – neurobiological mechanisms Impairments in motivated behaviour are a key feature in many stress-related disorders. Here, we investigate the effects of stress on motivated behaviour in rats differing in anxiety. We use a combination of techniques including, but not limited to, behavioural analyses (EPM, operant conditioning), microdialysis, immunofluorescent detection and quantification of protein expression, characterization of mitochondrial function and antisense-mediated modulation of gene expression. Keywords: Supervisor: Carmen Sandi Contact: [email protected] Required: – Availability: Open |
Neuroscience Wet |
| La Manno | Cerebellar development modeling in 4d Background: Current paradigms of studying the brain involve carrying out various kinds of measurements on tissue sections using different techniques. These measurements reveal different aspects of the brain complexity, such as its connectivity, electrophysiology properties, and anatomical relationships but also the gene expression, lineage relationship, and epigenetic states. These properties are all important and, in concert, give rise to complex emergent behavior. The advent of high throughput technologies has meant that data is being generated faster than ever. In this context, it becomes essential to integrate data from different modalities in an automated manner and study them in their spatial context to gain a fuller understanding of the brain and its function. Activities: The student will work to improve and refine methods to create a reference 3D atlas of a mouse brain for different ages. The volumetric reference generated will be used for the visualization and rendering of different kinds of data. Different kinds of omics and functional data will be integrated into this reference and correlated to generate new hypotheses and discoveries. In particular, we will focus on the cerebellum and Purkinje cells, that are an anatomically well-organized cell population, and it will facilitate the dynamical modeling of molecular changes through time during development. Keywords: Neuroscience, Data Science, Image analysis, Development Supervisor: Gioele La Manno Co-supervisor: Ludovic Telley (UNIL) Contact: [email protected] Required: Python or R programming. some experience with single processing or image analysis Availability: Open |
Computational Biology Interdisciplinary |
| Schürmann | Inference of ion-channel models using neural networks A central aspect of morphologically detailed neuron models is to accurately capture membrane mechanism dynamics. Ion channel models typically are described using a set of ordinary-differential equations and a set of parameters, which are fitted to experimental measurements. The parameter fitting process has traditionally been done using expert knowledge and parameter optimization methods, which find optimal parameters by comparing the electrical traces from experiment and simulation. With the advent of automated experimental methods for isolating and describing neuron ion channels, it becomes clear that a more automated, yet robust method for the generation of mechanism models, and their parameters, is needed. Neural networks may offer a new approach for accomplishing these tasks. Deep-learning approaches have been used in many instances for learning of parameters as well as the mathematical model itself (physics-informed neural networks). In this project we would like to explore in a first step the inference of optimal channel parameters by learning from the existing body of mechanism models using neural networks. Going further, the lessons learned could then be applied to designing neural networks for channel model inference. Keywords: deep neural networks, mathematical modelling, ion-channels Supervisor: Felix Schürmann Contact: [email protected] Required: machine learning, mathematical modelling, python programming, notions of membrane mechanism / ion-channel models Availability: Open |
Neuroscience Dry |
| Schürmann | Modeling energy-efficiency of neuron simulations Modelling and simulation of morphologically detailed neuronal circuits enables us to gain a deeper understanding of biological processes on multiple scales of the brain. In order to study complex phenomena such as neuronal plasticity we need to be able to run simulations at an increased scale, both in space and time. Understanding the various aspects of a neuron simulation that influence simulation performance and parallel scalability allows us to take decisions on hardware choice and improvements in algorithms and data structures, which are necessary to push the envelope of simulation scale. An important measure that has not been considered for this type of simulations is power consumption. In this project we would like to extend our performance models for neuroscience simulations with a model for power consumption. Existing power consumption models (e.g. based on ECM) should be explored and extended to fit our case. The developed model can then be incorporated into the performance model, allowing us to give more detailed insights into the expected feasibility and efficiency of simulations on various hardware platforms. Keywords: computational biology, simulation, mathematical modelling Supervisor: Felix Schürmann Contact: [email protected] Required: python programming, mathematical modelling, computer hardware architecture, basic understanding of morphologically detailed neuron simulations Availability: Open |
Neuroscience Dry |
| Blanke | Neural correlates of memory, navigation, and bodily self In this project, we will investigate how bodily self-consciousness influences episodic autobiographical memory and spatial navigation using Virtual Reality (VR) tasks. Two projects, using immersive virtual reality will be conducted. The first investigates the neural correlates of spatial navigation using EEG and the second the neural correlates of sensorimotor modulation of episodic autobiographical memory using fMRI. Keywords: memory, navigation, virtual reality, eeg, fmri, self Supervisor: Olaf Blanke Contact: [email protected] Required: – Availability: Open |
Neuroscience Dry |
| Blanke | Investigating bodily self-conciousness using virtual reality This project will combine virtual reality with neuroimaging and/or psychophysiology to explore the mechanisms underpinning the different dimensions of bodily self-conciousness, in healthy or clinical population. Keywords: multisensory integration, interoception, self-consciousness, virtual reality Supervisor: Olaf Blanke Contact: [email protected] Required: – Availability: Open |
Neuroscience Dry |
| Blanke | Vestibular contributions to bodily self-consciousness (motion platform) In this project you will combine virtual reality with the use of a motion platform in order to (1) induce virtual out-of-body experience in health subjects and (2) to explore the mechanisms underpinning such phenomenological experience. Keywords: vestibular processing, out-of-body experience, virtual reality Supervisor: Olaf Blanke Contact: [email protected] Required: – Availability: Open |
Neuroscience Dry |
| Blanke | Investigating hallucinations in Parkinson’s Disease The project will combine robotic technology, VR, psychophysics, and brain imaging technics to caracterize the brain mechanisms of hallucinations in PD. In particular, we aim at investigating how specific robotic sensorimotor stimulation can induce abberrant perceptions and modulate neural processes such as vision or audition. Keywords: parkinson’s disease, robotics, hallucinations, virtual reality Supervisor: Olaf Blanke Contact: [email protected] Required: – Availability: Open |
Neuroscience Dry |
| Mathis | Tackling animal cognition tests with RL to gain insights into brain function All animals possess some sort of intelligence, which manifests itself in more or less complex behavior, which typically allows them to solve specific tasks. To gain insights into brain function neuroscientists have devised a wide range of careful experiments to assess the potential and the limits of intelligence in different species, producing an extensive literature on the subject. Recently this literature inspired the Animal-AI testbed (http://animalaiolympics.com/AAI/), a framework to train and test autonomous agents to solve various classes of tasks inspired by the ones that animals face in laboratories. In 2019, tens of researchers have picked up this challenge, which focusses on training an autonomous agent to excel at as many animal-like skills as possible. Specifically, the agents are trained at solving typical laboratory problems, mostly related to acquiring food in various settings. However, not even the best submitted agents proved able to find a solution to the most complex tasks robustly. This confirms the suspicion that learning advanced animal behavior with Reinforcement Learning is still an open problem. The core of this thesis will be to tackle this challenge with the hope to push the envelope on one or more of the main forms of intelligence under examination in the testbed (spatial memory, dealing with transparent objects, object permanence, internal models, using tools and creativity). Keywords: RL, deep learning, behavior Supervisor: Alexander Mathis Contact: [email protected] Required: Python programming Availability: Open |
Neuroscience Dry |
| Suter | Protein homeostasis during early development How cells coordinate protein synthesis and degradation rates to regulate protein levels is poorly understood. We know very little on how different cell types vary in their strategies to regulate their proteome. Using a single cell, live approach that we pioneered (Alber et al., Molecular Cell 2018), this project will focus on how protein synthesis and degradation are modulated during embryonic stem (ES) cell differentiation, and how these depend on cell proliferation. The project will involve differentiation of various ES cell lines expressing a fluorescent timer, immunofluorescence analysis to determine the identity of differentiated cells, fluorescence microscopy, quantitative image analysis, and computational modelling of proteostasis. Keywords: developmental biology, stem cells, quantitative imaging, protein homeostasis Supervisor: David Suter Contact: [email protected] Required: None Availability: Open |
Cell biology Dry and wet |
| Suter | Quantitative analysis of transcription factor activity The activity of transcription factors (TFs) on cell fate decisions depends on their cumulative concentration over time. However monitoring their accumulation over time in single cells is challenging. The project will aim at developing a strategy allowing progressive changes in the expression levels of two spectrally distinct fluorescent proteins as a function of TF levels. The project will first provide a proof-of-concept of this strategy using in vitro experiments on mouse embryonic stem (ES) cells genetically modified with the reporter constructs. If time permits, genome editing will be performed to control the quantitative transcription factor readout by endogenous expression of Pax6 (a TF gene involved in neuronal cell fate determination). This will allow to track its cumulative expression during targeted in vitro differentiation of ES cells. This strategy will open the door to next generation, quantitative lineage tracing approaches and should have a broad range of applications to track the fate of cells with different expression levels of TFs or other genes in various organisms. Keywords: molecular biology, transcription factors, stem cells, quantitative imaging, Supervisor: David Suter Contact: [email protected] Required: – Availability: Open |
Cell biology Dry and wet |
| Turcatti | Classifying drugs according to their mode-of-action through High Content Screening High content analysis using automated fluorescence microscopy is widely applied in our lab for morphological profiling of cells upon interaction with drugs or candidate chemical compounds. Highly informative data extracted from fluorescence microscopy imaging of cells allows identifying and classifying a certain number of phenotypes clustered according to the biological signature obtained In order to increase the level of information related to intracellular events triggered by chemical interference, the morphological profiling assay ‘cell painting’ will be implemented. In this multiplexed method, six fluorescent probes are used simultaneously for revealing cellular compartments or organelles under chemical perturbation. By automated image analysis and extraction of hundreds of features, and using machine learning algorithms, compounds will be clustered according to the phenotypic profile they trigger. For this multidisciplinary chemical biology project, experiments will initially be performed with one cell line at a fixed time point and at a single concentration. For deeper characterization and for few selected compounds, fluorescence-imaging time-lapse experiments can be envisioned for tracking the phenotypic evolution over time. Open source supervised machine learning software and solutions such as Cell profiler or Cell cognition will be used for image analysis and throughout the project. Keywords: Chemical Biology, Drug Discovery, Phenotypic screening, Image analysis Supervisor: Gerardo Turcatti Contact: [email protected] Required: Some experience in mammalian cell culture and/or fluorescence microscopy Availability: Open |
Bioengineering Dry and wet |
| Jakšić | Cognitive fitness from a genetic, metabolic, and evolutionary perspective The risk for developing Alzheimer’s disease (AD) is inversely correlated with the degree of educational attainment: The higher your educational achievement, the lower your risk for AD. This association is called the “cognitive reserve hypothesis” and, although epidemiologically supported by multiple studies, the physiological and molecular mechanisms behind this association are poorly understood. In this project, we set out to test the cognitive reserve hypothesis from an energy demand perspective. Since both learning and neurotoxicity in AD pose a high demand on cellular energy consumption, we hypothesize that well-trained brains are conditioned to utilize energy more economically. In other words, energy saved by a well-trained brain on learning can be re-routed to protect against neuronal cell death. To experimentally assess this hypothesis, you will use the fruitfly D. melanogaster we well as murine models of AD in a joint project between the Jaksic and Gräff labs. Specifically, you will first assess the genetic basis for the relationship between energy consumption, learning-dependent metabolic rates, lifetime learning, and neurodegeneration in a screen of genetically characterized D. melanogaster lines. These results will then be validated by whole transcriptome sequencing and functional manipulations in different mouse models of AD. Knowing the genetic architecture underlying metabolism, lifetime learning and the risk for AD will yield important insights into molecular basis of the cognitive reserve hypothesis. Interested? Please contact us at [email protected] or [email protected]. Keywords: Genomics, Neurobiology, Bioinformatics, Cognition Supervisor: Ana Jaksic Co-supervisor: Gräff Contact: [email protected] Required: R and/or Python, willingness to learn to work with the fly model Availability: Open |
Neuroscience Interdisciplinary |
| Aztekin | Limb regeneration associated cell type cultures Unlike mammals, certain species can regrow their lost limbs. Limb regeneration involves complex cell-cell interactions mediating proliferation, migration, and cell-fate decisions. Notably, upon limb amputations, Xenopus laevis tadpoles form an epithelial cell type (Apical-ectodermal-ridge, AER cells) that expresses genes influencing stem and progenitor cells for cellular mechanisms to regrow lost limbs. However, how AER coordinates multiple dynamic behaviors of stem and progenitor cells for morphogenesis is unclear. To overcome this, we will establish an in vitro system to investigate AER cells and their interaction with other populations, where imaging and molecular biology approaches could be readily employed. The student will focus on isolating and establishing in vitro culture protocol for AER cells. Upon successfully propagating these cells, the student will implement high-throughput live-imaging, gene-editing, and co-culture studies (AER cells with stem/progenitor cell types) to interrogate cell-cell interactions and dynamic cellular behaviors. Keywords: regeneration, stem cells, signaling centers, AER, development, limb Supervisor: Can Aztekin Contact: [email protected] Required: Experience in cell culture Availability: Open |
Developmental biology Wet |
| Deplancke | Studying the genetic basis of variation in mosquito vectorial capacity in Rio de Janeiro Dengue and chikungunya are mosquito (Aedes aegypti)-borne arboviral diseases that affect millions of people, resulting in substantial hospitalization and death. To combat these diseases, we need to achieve a deeper understanding of A. aegypti biology and the impact of viral infection on the life-history traits of this insect, as this should have important implications for predicting the evolution of mosquito-parasite relationships and their role in the emergence and maintenance of arboviruses. To achieve such understanding, the Laboratories of Systems Biology and Genetics (EPFL) and Physiology and Control of Vector Arthropods (FIOCRUZ, Rio de Janeiro) have already generated a panel of already >60 recombinant A. aegypti inbred lines. We have begun to systematically sequence the genomes of these lines and we are now looking for a Master’s student who will engage in bioinformatic analyses to map genomic variation across these >60 lines. In parallel, the student will work in a fully equipped insectary at Fiocruz, performing phenotyping experiments on these inbred lines including the number of eggs deposited by females, egg viability, adult longevity, larval development time, insecticide resistance etc. Finally, the student may engage in some first genotype-phenotype analyses, linking specific mosquito traits to genetic variants. Keywords: Mosquito, infection, vectorial capacity, viral disease, genomic variation, systems genetics, bioinformatics Supervisor: Bart Deplancke Co-supervisor: FIOCRUZ, Rio de Janeiro Contact: [email protected] Required: Skills in R / Python Availability: Open |
Infectious diseases Interdisciplinary |
| Dal Peraro | Simulating Nanopore Transport as a Heteropolymer Sequencing Technique Nanopores are molecular assemblies that form cylindrical holes in a cell membrane or artificial surface. When a heteropolymer such as a protein translocates through a nanopore, the transient blockage of the channel depends on the sequence of monomers and can be quantified by the variation in an ionic current through the pore. The time series of the current provides a readout of the polymer’s sequence. The complex interactions of the polymer with the pore’s inner surface and the slow (millisecond) timescale of the process make experimental optimisation of nanopore behaviour slow and expensive. Molecular simulations provide near-atomistic detail on the length scale of amino acids, and can reach time scales of hundreds of microseconds. This project will use the established technique of dissipative particle dynamics simulations to simulate a heteropolymer translocating through a nanopore in a lipid membrane. The monomer-dependent variation in occlusion will be read out and its relation to the sequence of the polymer determined. The influence of pore parameters — width, surface structure — and a range of polymer sequences will be explored. The results are expected to shed light on the sensitivity and precision with which a polymer’s identity can be determined from its translocation through the pore. Keywords: nanopore, aerolysin, sequence, membrane, simulation Supervisor: Matteo Dal Peraro Contact: [email protected] Required: Good familiarity with molecular simulations; knowledge of C++ a bonus Availability: Open |
Bioengineering Dry |
| Dal Peraro | Developing biological nanopores for molecular sensing Evolution has found countless ways to transport material across cells and cellular compartments separated by membranes. Protein assemblies that form channels and pores regulate the passage of molecules in and out of cells, contributing to maintain most of the fundamental processes that sustain living organisms. We are taking advantage of the natural properties of these biological systems to push technology forward and to develop the single-molecule sensing devices of tomorrow. Since we solved the structure of aerolysin, we are working to characterize and engineer this and other similar pores in order to enhance their native molecular sensing capabilities. Current fields of application that we are exploring in the lab include protein sequencing, biomarkers identification, polymer reading for long-term data storage and detection of environmental chemicals. Moreover, we strive to improve the current instrumentation and the signal processing pipeline using artificial intelligence. Multiple master projects are available in this domain both on the experimental or computational side depending on the skills and inclination of the candidate. Keywords: nanopore sensing, protein sequencing, protein engineering, data storage Supervisor: Matteo Dal Peraro Co-supervisor: Chan Cao, SV Contact: [email protected] Required: Some of these are preferable : protein production, nanopore sensing, signal processing, structural biology Availability: Open |
Bioengineering Interdisciplinary |
| CIBM MRI EPFL AIT | TP4 and Master projects @ CIBM MRI EPFL AIT CIBM is a Swiss Research Centre of Excellence in biomedical imaging. CIBM brings together highly qualified, diverse, complementary and multidisciplinary groups of people with common interest in biomedical imaging. Several projects in the field of Magnetic Resonance Spectroscopy and Imaging applied to study brain metabolism, microstructure and function in health and disease are available https://cibm.ch/open-positions/ Keywords: Bioengineering, Neuroscience, Brain metabolism, Magnetic Resonance Spectroscopy and Imaging, microstructure Supervisor: Cristina Cudalbu Co-supervisor: [email protected] Contact: [email protected] Required: – Availability: Open |
Bioengineering Interdisciplinary |
| Schürmann | Simulation of electrical neurostimulation Electrical stimulation is used to evoke neural activity in the cortex for both experimental and clinical purposes, but it’s mechanism of action, especially at the circuit level, is not well understood. In silico experiments can be used to disentangle the direct effect of the applied electric field from that of the propagated activity, and to optimize electrical stimulation parameters for efficacy and selectivity. In this project, you will use the Blue Brain Project’s reconstruction of rat cortical microcircuitry, as well as the IT’IS Foundation’s finite element models of the rat head, in order to understand the synaptic pathways that govern the propagation of electrically-triggered activity, and to understand the influence of tissue and electrode geometry on the evoked neural activity. Keywords: neurostimulation, computational neuroscience, finite element models. Supervisor: Michael Reimann Co-supervisor: Michael Reimann (BBP) Contact: [email protected] Required: Python, basic knowledge of computational neuroscience and electromagnetics. Experience with Linux, bash scripting, and/or finite element modeling is a plus Availability: Open |
Neuroscience Interdisciplinary |
| Duboule | Development of ESCAPE: the Extended Single Cell Atlas of Pseudo Embryos This master project aims at producing a single cell atlas comparing different models of pseudo-embryos. These stem cell-derived models of embryos are generating a lot of excitement in the field of developmental biology as they hold the promise to study embryos inside a dish. With minimal input, these are able to self-organize into structures that resemble the embryos. Recent years have seen an explosion of these models and each are tailored to study specific embryonic stages or systems. However, no systematic comparison of these models has been done and it is unclear as of yet how each compare to the other. As such, it can be challenging to identify which model is best tailored to one’s research studies. This master project aims at filling that gap by integrating several datasets of single cell RNA seq with embryonic datasets in order to comprehensively assess what each model is best at. To perform this project, the student will have access to the High-Performance Computing server of EPFL as well as a lab private Keywords: single-cell transcriptomics, bioinformatics, data analysis Supervisor: Denis Duboule Co-supervisor: [email protected] Contact: [email protected] Required: coding (Unix, R, and/or Python). Single cell data analysis is a strong plus Availability: Open |
Computational Biology Interdisciplinary |
| Duboule | Functional analysis of pseudo-embryo mutant This master project will use pseudo-embryos to investigate the phenotypic impacts of different genetic mutations. The master student will generate gastruloids and other models of embryos using embryonic stem cells that have been genetically modified to inactivate several genes. Several genetic mutations will be evaluated at a morphological and functional level. Immuno-fluorescence, and RNA-seq and other functional assays will be performed to evaluate the defaults in development of these pseudo-embryos. Keywords: genetic, organoids, CRISPR/Cas9, image analysis Supervisor: Denis Duboule Co-supervisor: [email protected] Contact: [email protected] Required: Useful skills: Microscopy, image analysis, cell culture, molecular biology Availability: Open |
Developmental biology Interdisciplinary |
| Courtine | A computational approach to optimize spinal nerve recruitment during electrical stimulation Spinal cord injury disrupts the communication between the brain and the spinal circuits below the lesion that generate and coordinate limb movements. Epidural Electrical Stimulation (EES) can restore movements of paralyzed limbs. To effectively restore walking, therapists need to calibrate a spectrum of stimulation protocols. Even for a simple functional movement such as walking, the calibration is performed over several weeks by expert engineers. This process presents a barrier for real-life worldwide use of the therapy. In this project, we propose to optimize spinal nerve recruitment during EES. The master student will work on optimizing the electrical field potential resulting from a multipolar electrode configuration. To do so, the student will use our computational model of the patient’s spine. In order to push further results established by our optimizer, the student must be familiar with concepts of Computational Neuroscience, Artificial Intelligence and programming in Python/Pytorch. To predict the resulting motor pattern, the student will also work with data such as recorded muscle activities and/or kinematic data. The student will develop a computational framework to obtain the proprioceptive map to better understand the link between electrode configuration and muscle recruitment. The developed method will be validated using clinical data. Keywords: Computational Neuroscience, Finite Element Model, Spinal Cord, Artificial Intelligence Supervisor: Gregory Dumont Contact: [email protected] Required: Python – Matplotlib/Mayavi/Pytorch/Pandas – Artificial intelligence/Machine learning/Data analysis – Computational Neuroscience Availability: Open |
Computational Biology Dry |
| Lingner | telomere biology The projects are linked to our ongoing research work on telomeres, involving molecular biological, biochemical and cell biological techniques and human cell culture. See https://www.epfl.ch/labs/lingner-lab/research-activities/ and publications. Keywords: molecular biology, cell biology, cancer biology, biochemistry Supervisor: Joachim Lingner Contact: [email protected] Required: strong motivation Availability: Open |
Molecular biology Wet |
| Herzog | Integrative psychophysical benchmarking of deep neural network models of the visual stream on Brain-Score Neural benchmarking of deep neural networks of the ventral stream has produced models with improved explanatory power of primate neural data. However, issues remain: neural benchmarks fail to consistently discriminate between models of vastly different architecture and training methodology. Furthermore, deep neural networks in general are unable to explain myriad psychophysical results observed in humans, and unified benchmarks of psychophysics in deep neural networks do not yet exist. Implementing such benchmarks is important, because a wide body of literature on psychophysical tests of deep neural networks shows that such tests able to discriminate between deep neural network models and inform future model-building. In this project, the student will develop and organize human experiments inspired by the vast body of work from psychology, and that are explicitly tailored to behaviorally compare specific computations in deep neural networks and humans. The student will then test and implement these experiments as behavioral metrics on Brain-Score. Keywords: deep neural networks, psychophysics, vision Supervisor: Michael Herzog Contact: [email protected] Required: Programming – high proficiency in python; familiarity with any of MATLAB, PyTorch, Keras or Tensorflow are a plus but not required Availability: Open |
Neuroscience Dry |
| Ramdya | Apply deep learning to study animal behavior Insects generate complex behaviors with a numerically simple nervous system. The objective of this project is to leverage computer vision and deep learning algorithms to study the leg positions of tethered behaving flies from high-resolution movies. These behavioral sequences can then be linked to simultaneously acquired neuroimaging data. This project at the interface of computer science, and neurobiology will be supervised at the Neuroengineering Laboratory in close interaction with computer science groups on campus. Keywords: Supervisor: Pavan Ramdya Co-supervisor: Fua (IC) Contact: [email protected] Required: C/C++, and/or Python Availability: Open |
Neuroscience Interdisciplinary |
| Ramdya | Build a neuromechanical insect simulation To understand the behavior of complex systems it is often necessary to build a model. The goal of this project is to develop a biorealistic 3D simulation of Drosophila. This model will be used to test bioinspired neural networks limb controllers. The project at the interface between robotics, computer science, and neurobiology will be supervised at the Neuroengineering Laboratory in close collaboration with the Biorobotics Laboratory. Keywords: Supervisor: Pavan Ramdya Co-supervisor: Ijspeert (Robotics) Contact: [email protected] Required: C/C++, and/or Python Availability: Open |
Neuroscience Interdisciplinary |
| Ramdya | Apply deep learning to study neural circuits A central goal of neuroscience is to link neural activity and behavior. The objective of this project is to use computer vision and deep learning approaches to extract neural activity patterns during behavior and to make accurate predictions behavioral and internal states. This project at the interface between computer science and neurobiology will be supervised at the Neuroengineering Laboratory in close interaction with computer science and image processing groups on campus. Keywords: Supervisor: Pavan Ramdya Co-supervisor: Fua (IC) Contact: [email protected] Required: C/C++, and/or Python Availability: Open |
Neuroscience Interdisciplinary |
| Ramdya | Build a robotic fly Nature has solved numerous challenges associated with autonomous behavioral control. We hope to leverage these solutions in robotics. The goal of this project is to construct an insect-inspired robot and to test bioinspired algorithms of limb control. This project at the interface of robotics and biology will be supervised at the Neuroengineering Laboratory in collaboration with EPFL robotics groups. Keywords: Supervisor: Pavan Ramdya Co-supervisor: Ijspeert (Robotics) Contact: [email protected] Required: Microfabrication and/or Electronics experience Availability: Open |
Neuroscience Interdisciplinary |
| Ramdya | Build robotic systems to automate neuroscience microsurgery Surgical interventions commonly performed in medicine and research require skill and extensive training. The objective of this project is to democratize and automate microsurgery by developing automated vision-guided robotic systems that dissect and prepare animals for neural recordings. This project at the interface between robotics and neurobiology will be supervised at the Neuroengineering Laboratory in collaboration with the Microrobotics Laboratory. Keywords: Supervisor: Pavan Ramdya Co-supervisor: Sakar (Robotics) Contact: [email protected] Required: Robotics / electronics experience Availability: Open |
Bioengineering Interdisciplinary |
| Bitbol | Exploring the sequence-function relationship in proteins Proteins play crucial roles in our cells. The amino-acid sequence of a protein encodes its function, including its structure and its possible interactions. In evolution, random mutations affect the sequence, while natural selection acts at the level of function. Shedding light on the sequence-function mapping of proteins is central to a systems-level understanding of cells, and has far-reaching applications in synthetic biology and drug targeting. The current explosion of available sequence data enables data-driven approaches to discover the principles of protein operation. In alignments of homologous protein sequences, correlations exist between certain amino-acid sites. They can arise from functional optimization but also from evolutionary history, and disentangling these two types of correlations is an important challenge. We employ methods based on information theory, statistical inference and machine learning to investigate this. Several directions are possible, but the project will be data-driven and involve analyzing real protein sequence data and/or protein structure data. Keywords: protein sequences, computational biology, protein function, evolution Supervisor: Anne-florence Bitbol Contact: [email protected] Required: – Availability: Open |
Computational Biology Dry |
| Mathis | Task-driven hierarchical deep neural networks model of the sensorimotor pathway Proprioception and touch are critical for controlling the body and motor control, yet no canonical hierarchical model of the system exists. Task-driven modeling has provided important insights into other sensory systems. To study proprioception and touch we design different tasks as well as train models of the sensorimotor pathway on various behavioral tasks. We will then analyze the networks’ units as well as emerging computations to gain insights into proprioception, touch and motor control. We are specifically interested in understanding the implications of different tasks, movement statistics and network architectures. Check out https://www.mathislab.org for more details and related literature. Keywords: task-driven modeling, deep learning, proprioception, touch, motor control, reinforcement learning Supervisor: Alexander Mathis Contact: [email protected] Required: Python programming skills • Knowledge in TensorFlow/PyTorch would be great (not required) Availability: Open |
Computational Biology Dry |
| Mathis | Behavioral analysis with deep learning We strive to develop tools for the analysis of animal behavior. Behavior is a complex reflexion of an animal’s goals, state and individuality. Thus, accurately measuring behavior is crucial for advancing basic neuroscience, as well as the study of various neural and psychiatric disorders. The goal of the master thesis project is to develop and optimize deep neural network architectures in order to predict the behavioral state of animals in various experiments with limited amounts of data. Depending on the interest of the student, appropriate datasets ranging from analysis of neural disorders in humans to social behavior in mice will be used. Check out https://www.mathislab.org for more details and related literature. Keywords: DeepLabCut, action recognition, deep learning, behavior Supervisor: Alexander Mathis Contact: [email protected] Required: Python programming skills • Knowledge in TensorFlow/PyTorch would be great (not required) Availability: Open |
Computational Biology Dry |
| Mathis | Assessing human eating behavior with deep learning A joint master internship is available in the Mathis Group at EPFL and the Food Cognition Group at the Institut Paul Bocuse research center (IPBR) in Lyon to measure human eating behavior from video-recordings using machine-learning and computer vision approaches. This position is not funded but travels to visit IPBR in France will be covered. Project overview: Eating is a daily and essential activity that implicitly encompasses a lot information. Our eating behavior is a result of our physiological needs, past experiences, as well as social and cultural environment, but it can also reflect our health status. Despite evidence linking changes in eating habits and the development of psychiatric disorders, we are still lacking a robust and fine-grained measures that can reliably detect the early signs of these diseases. The aim of this project is to advance this search using naturalistic video-recordings of healthy participants eating at IPBR restaurants. This research will leverage advanced machine-learning analyses and computer vision techniques to identify actions, as well as implicit information during eating. Keywords: DeepLabCut, action recognition, deep learning, behavior Supervisor: Alexander Mathis Contact: [email protected] Required: Python programming skills • Knowledge in TensorFlow/PyTorch would be great (not required) Availability: Open |
Neuroscience Dry |
| D’Angelo | Regulation of Lipid metabolism in Cell Growth While proliferating, cells need to duplicate not just their DNA but also their cytosolic and membrane components. Cell growth is indeed the outcome of multiple processes, including the de novo synthesis of proteins and lipids. To this purpose, during phases of active cell division, the rate of growth is matched to the availability of nutrients required for the coordinated protein biosynthesis and membrane expansion. How the lipid synthetic system is regulated in response to growth stimuli or nutrients availability and how lipid remodelling influences growth is partially understood. Here we will evaluate changes in lipid metabolism as a consequence of different growth conditions and the impact of lipid synthetic inhibition to cell growth. Keywords: Lipid metabolism, proliferation and cancer, intracellular trafficking Supervisor: Giovanni D’Angelo Contact: [email protected] Required: – Availability: Open |
Cell biology Wet |
| Fellay | Leveraging genomics to understand host-pathogen interactions Hosts and pathogens are in a constant battle characterized by successive rounds of evolution. To understand such a process, paired host and pathogen genomes from infected patients can be sequenced. Genetic variations that co-occur can then be pinpointed; elucidating mechanisms involved in in host-pathogen interactions. Depending on interest, two possible projects are proposed: i) The student will create a bioinformatics pipeline (using the workflow management software, Snakemake) such that the abovementioned approach could be applied by the wider scientific community. An artificial dataset will be created through simulations to ensure the validity of the pipeline. ii) In the context of the pathogen Mycobacterium tuberculosis, the student will work on network-based approaches that leverage paired host-bacterial genome sequences to elucidate host and pathogen pathways relevant in Tuberculosis. Keywords: host-pathogen interactions, genomics, bioinformatics, infectious diseases Supervisor: Jacques Fellay Contact: [email protected] Required: General programming knowledge (R and/or Python). Knowledge in bioinformatics and statistics. Experience with Unix command line and BASH script beneficial. Availability: Open |
Computational Biology Dry |
| Fellay | Adaptation of ACMG/AMP variant classification framework for primary immunodeficiency associated genes Primary immunodeficiencies (PIDs) are rare genetic disorders that impair the immune system. Without a fully functional immune response, people with PID may develop severe symptoms upon infection with common pathogens such as respiratory viruses. To find the genetic cause of PIDs, we use a genome sequencing approach, which yields thousands of candidate variants per individual. The ACMG/AMP classification is used to identify the most likely deleterious mutations and prioritize them for downstream functional analysis. However, the ACMG/AMP classification was developed for highly penetrant Mendelian disorders and needs to be adapted for immune-related diseases, which are more likely to be subjected to balancing selection. This project aims at adapting the ACMG/AMP classification for PIDs by investigating multiple aspects of immune-related genes such as population allele frequency, genetic intolerance to mutation, and mutational hotspots. It will contribute to a deeper understanding of the genetic causes of immunodeficiencies, which could open up new avenues for diagnostic and therapeutic development. Keywords: bioinformatics, statistics, immunity Supervisor: Jacques Fellay Contact: [email protected] Required: Experience with programming languages R, python, or unix shell scripting (bash) are beneficial for data manipulation. A biological sciences background will benefit the interpretation of biological-relevance Availability: Open |
Computational Biology Dry |
| Cao | Structure–function relationship of pore-forming toxins and their application in molecular sensing Pore forming toxins (PFTs), an ancient protein family, produced by many pathogenic bacteria have long fascinated structural biologists, microbiologists and immunologists. Recently, this family of proteins is growing attention also because of their biotechnological application as nanopore sensors for biological and synthetic molecules sensing and sequencing. During this project, by engineering single and/or multi amino acids of aerolysin, we expect to tune the sensing capability of aerolysin nanopores for biomolecules of interested, and therefore achieving a more accurate molecular detection. This will have impact on nanopore single-molecule DNA/Protein sequencing. Keywords: Bioengineering, biophysics, biosensor, single-molecule detection Supervisor: Chan Cao Contact: [email protected] Required: – Availability: Open |
Bioengineering Wet |
| Cao | First steps toward single-molecule sensing of protein biomarkers by using novel bioengineered nanopores We are looking for some students interested in the development of nanopore-based next generation DNA and protein sensing and sequencing approaches. Biological nanopores have been developed mainly from pore-forming toxins and used for the single-molecule detection of DNA, small peptides, and other small molecules, with already some commercial applications. In this project, the student will test and assess several novel biologically engineered nanopores (mutants or chimeras of existing biological nanopores), which have been designed to improve the sensitivity and efficiency towards peptides detection. Specifically, we are interested in the detection of unique protein biomarkers for neurodegenerative diseases from heterogeneous mixture samples, but also in the development of more sensitive biological nanopore to better assess peptides composition. This work is the first step towards single-molecule protein sensing and sequencing, which is probably the next revolution in biology, since it could help us to understand many fundamental biological processes and help in the early discovery of peptides biomarkers for nowadays incurable diseases, such as Parkinson or Alzheimer disease. Keywords: Nanopore, protein biomarkers, peptides, neurodegenerative diseases Supervisor: Chan Cao Contact: [email protected] Required: – Availability: Open |
Bioengineering Wet |
| Cao | Nanopore-based single-molecule proteomics Biological nanopores have been successfully applied in sequencing long DNA, and this achievement has inspired its application for single-molecule proteomics, which is an exciting and emerging field to explore. The project will focus on applying various biological pores (that are available in the lab) for application in single-molecule proteomics, including protein sensing and/or sequencing. Keywords: Nanopore, biophysics, protein engineering Supervisor: Chan Cao Contact: [email protected] Required: – Availability: Open |
Bioengineering Wet |
| Herzog | Temporal integration of visual information at multiple time scales Perception involves the integration of sensory information at multiple time scales. In our research, we study how the visual system combines information over milliseconds and seconds to construct our conscious experience of the world. We use a multidisciplinary approach, combining human psychophysics, computational modeling, electroencephalography (EEG) and functional magnetic resonance imaging (fMRI). Our final objective is to characterize the behavioral and neural correlates of the mechanisms that support the temporal integration of visual events at multilple level of processing, from perception to memory and decision-making. The student will have the opportunity to focus on one specific techique (e.g., psychophysics, EEG or fMRI) or a combination of techniques, and will be involved in the collection, analysis and modeling of different types of data. Keywords: vision, computational modeling, EEG, fMRI Supervisor: David Pascucci Co-supervisor: Michael Herzog, SV Contact: [email protected] Required: Good analytical and computational skills, ideal for students in: Computer Sciences, Life Sciences, Bioengeneering Availability: Open |
Neuroscience Interdisciplinary |
| Fellay | Mutational spectra of RNA viruses: uncovering the species-specific endogenous and exogenous mutagens The mutational spectrum, defined as the types and frequencies of nucleotide substitutions at neutral positions, reflects a set of mutagens affecting a given genome. For example, the mutational processes at play in various types of human cancers are known, which allows the deconvolution of the observed mutational spectrum into signatures of specific mutagens, such as UV light, tobacco smoke, etc (https://cancer.sanger.ac.uk/signatures/). However, progress in the understanding of the mutational processes of different species is much more modest. In this project we plan to reconstruct the mutational spectrum for different RNA viruses (i.e., download sequences, align them, build phylogenetic trees, reconstruct all mutations) and link it to viral life-cycle peculiarities in order to uncover species-specific mutagens. Finally, we will be able to answer several interesting questions: (i) whether the mutational spectra of SARS-CoV-2, Zika, Ebola, Influenza, etc. are different? why? (ii) whether the mutational spectrum of SARS-CoV-2 depends on patient age, sex or severity of disease? (iii) which mutagens (exo- or endogenous) contribute more to the mutational burden of various viral species? Keywords: mutational spectrum, RNA viruses, mutagens Supervisor: Jacques Fellay Contact: [email protected] Required: programming skills (python, R) Availability: Open |
Computational Biology Dry |
| Gönczy | Re-engineering SAS-6 proteins Objective: test if centrioles organized by engineered SAS-6 proteins with altered sizes and/or symmetries can form and function in human cells. Approaches: human cell culture, CRISPR/Cas9-mediated engineering, expansion microscopy, super-resolution microscopy. Keywords: cell biology, centriole organelle, CRISPR/Cas9, microscopy Supervisor: Pierre Gönczy Contact: [email protected] Required: Ideal for students in: Life Sciences, Bioengineering. Availability: Open |
Cell biology Wet |
| Gönczy | Analyzing centriole fate during zebrafish muscle formation Objective: monitor centrosome and centriole fate during muscle formation in zebrafish embryos Approaches: 4D live imaging using light-sheet microscopy, injection of RNA/DNA into zebrafish embryos, develop and apply tracking algorithms to monitor centrosomes and centrioles. Keywords: cell and developmental biology, muscle formation, zebrafish embryos, centrosome and centriole, live imaging Supervisor: Pierre Gönczy Co-supervisor: Oates Contact: [email protected] Required: Ideal for students in: Life Sciences, Bioengineering. Availability: Open |
Developmental biology Interdisciplinary |
| Gönczy | Analyzing novel centriolar proteins in Chlamydomonas reinhardtii Objective: identify the localization and test the function of novel centriolar proteins in the green alga Chlamydomonas reinhardtii. Approaches: CRISPR/Cas9-mediated GFP tagging, as well as disruption, of novel centriolar proteins in Chlamydomonas reinhardtii, expansion microscopy, super-resolution microscopy. Keywords: cell biology, green algae, centriole, CRISPR/Cas9, microscopy Supervisor: Pierre Gönczy Contact: [email protected] Required: Ideal for students in: Life Sciences, Bioengineering. Availability: Open |
Cell biology Wet |
| Gönczy | Mechanisms of centriole elimination during C. elegans embryogenesis Objective: discover genes that modulate centriole number using C. elegans embryos as model system. Approaches: RNAi-based functional genomic screen, live imaging, image analysis, molecular biology, cell biology. Keywords: cell and developmental biology, C. elegans, embryogenesis, functional genomics, microscopy Supervisor: Pierre Gönczy Contact: [email protected] Required: Ideal for students in: Life Sciences, Bioengineering. Availability: Open |
Developmental biology Wet |
| Gönczy | Mechanisms of symmetry breaking in the C. elegans zygote Objective: decipher mechanisms of symmetry breaking at the onset of development in the C. elegans embryo, using a combination of experimental and theoretical approaches. Approaches: molecular biology, CRISPR/Cas9-mediated engineering, live imaging, image analysis, mathematical modeling. Keywords: cell and developmental biology, C. elegans, embryogenesis, microscopy, mathematical modeling Supervisor: Pierre Gönczy Contact: [email protected] Required: Ideal for students in: Life Sciences, Bioengineering, Physics Availability: Open |
Developmental biology Dry and wet |
| Rahi | Experimental evolution of cell-cycle speed Cells divide at vastly different speeds, ranging from 10 min doubling time in the fastest growing bacterium to 90 min in budding yeast and around 24 h in human cell cultures. While single-cell organisms have presumably been optimized for dividing quickly and accurately during the course of evolution, it is not clear what the speed limits on growth and division are. This is a fundamental open question that would also be useful to answer for a variety of applications. We have engineered a budding yeast strain in which progress through the cell cycle is fully controlled using light. We are performing long-term laboratory evolution experiments in which we ‘force’ cells to divide faster. By combining the analysis of the cell cycle dynamics with whole genome sequencing of the evolved strains we want to uncover the gene variants that can make cells speed up. The thesis will include an experimental part (constructing budding yeast strains and running evolution and microscopy experiments) as well as a computational part where the cell lineages will be reconstructed from microscopy data and gene variants identified using whole genome sequencing data. The ideal candidate should be familiar with standard molecular biology techniques (PCR, cloning, bacterial transformation). Keywords: experimental evolution, cell-cycle speed, budding yeast, whole-genome sequencing Supervisor: Sahand Rahi Contact: [email protected] Required: Standard molecular biology techniques Availability: Open |
Cell biology Dry and wet |
| Pardon | 3D printed stamps for 3D cell culture: characterization and high-throughput implementation 3D modelling and test of a high-resolution stamp to microstructure hydrogels for organoids culture. This project aims to design stamps with CAD software and print it using a high-resolution stereolithography 3D printer. The 3D design and the stamp material will be validated by testing the microstructuring of various synthetic and natural hydrogels to evaluate the formation of organoids. The resulting microstructured hydrogel should enable to seed and culture organoids, should allow for stable entrapment inside the microstructures during automation processes, such as under rapid microscope stage movement and displacement with a robotic arm, or media exchange. Subsequently, the stamps and the hydrogel formation should be adapted for various cell culture well plates of size 6-12-48-96(-384), so that they can be used for higher-throughput experiments with an automated liquid handler. The effect of hydrogel stiffness and microwell shape on the morphogenesis of organoids and co-culture assay will also be investigated. This work will greatly contribute to the automation of organoid culture and analysis for the scientific community of the EPFL, CHUV, UNIL, UNIGE, and HUG. The contribution of the Master’s Student will help to solidify the way toward the establishment of a ‘live biobank’ of patient-derived organoids. Keywords: cad design, 3d printing, material/biomaterials, organoids, automation application, high-throughput. Supervisor: Gaspard Pardon Contact: [email protected] Required: CAD skills, basic cell culture and microscopy. Availability: Open |
Bioengineering Wet |
| Pardon | Development of an automated tissue clearings workflow for high-throughput 3D characterization of cancer organoid models. An organoid is a three-dimensional multicellular structure that shows realistic micro-anatomy of an organ. This in vitro model mimics the in vivo environment, architecture and multi-lineage differentiation of the original organs and allows to answer many interesting biological questions. For these reasons, they are widely used in stem cell, regenerative medicine, toxicology, pharmacology, and host-microbe interactions research. In order to study organoids, microscopy is very useful: It is possible to make three-dimensional reconstruction of serial sections but it is time consuming and error-prone. Here we propose an alternative solution: Tissue clearing reduces the dispersion of light because it homogenizes the refractive index of the tissue, allowing sample observation throughout its thickness.Therefore, the aim of this course is to compare different clearing techniques on organoids and spheroids and to automate the cleaning with the Plate Washer-Dispenser. The internship will be divided into 4 parts: 3D culture of different types of spheroids (colorectal cancer / glioblastoma) – Development of different Clearings tissue protocols (2ECI Ethyl Cinnamate / Cubic) depending on the organoid type – Implementation of different Clearing protocols on the EL406 wash dispenser for high throughput automation. – High throughput 3D imaging on Celldiscoverer 7 (Zeiss) and analysis of results. Keywords: Clearings tissue – 3D culture – organoid – automation – confocal microscopy Supervisor: Gaspard Pardon Co-supervisor: [email protected] Contact: Gaspard.pardon.epfl.ch Required: cell culture – IF – automation -microscopy Availability: Open |
Bioengineering Interdisciplinary |
| Herzog | Unifying peripheral vision with a deep neural network The perception of peripheral information in human vision is not well-understood – peripheral vision loss is not simply blur. While modern deep neural networks capture substantial parts of the visual stream’s capability to detect objects, no consensus on the critical question of how information is compressed in the periphery has been reached. Many seemingly distinct phenomena have ubiquitously been observed and characterized in humans, such as crowding, peripheral visual illusions, and pop-out in visual search. In this project, the student will investigate unifying these separate phenomena as loss optimization in a peripherally constrained deep neural network. The student will be involved in collecting and analyzing psychophysical data, and will implement a peripherally constrained deep neural network with the goal of simultaneously explaining these separate phenomena under one framework. The student will ideally either be proficient at deep learning and interested in human psychophysics and data analysis, or vice versa. Keywords: deep neural networks, psychophysics, peripheral vision, visual illusions Supervisor: Michael Herzog Contact: [email protected] Required: deep learning OR psychophysics & data analysis; familiarity with Python/PyTorch and MATLAB are a plus Availability: Open |
Neuroscience Dry and wet |
| Pardon | Machine learning and imaging-based feedback for ‘smart’ cancer organoids culture automation We are looking for a student to join our interdisciplinary team to develop ‘smart’ imaging algorithms, implemented in Python, as a key step towards the automated handling and the biobanking of patient-derived cancer organoid cultures. Patient-specific cancer organoid models hold great promise for the advancement of precision oncology. However, the currently employed manual organoid culturing techniques are cumbersome, tedious, and extremely time-consuming, thus greatly restricting the use, and further scale-up of patient-derived organoid cultures for large-scale drug screening towards personalized medicine applications. Seeking to realise the immense potential of patient-specific organoids, and paving the way towards patient-centred diagnostics and therapy, we aim to establish capabilities for the automated handling and expansion of cancer organoid cultures. Developing the ability to culture organoids in a fully automated manner, and, critically, under defined and highly reproducible conditions, requires identification of, and ‘smart’ feedback on critical features and parameters, such as organoid morphology, or culture density, in order to maintain long-term viability, and to obtain robust and homogenous experimental models. In this cutting-edge project, you will develop metrics and algorithms for computational image analysis and machine learning, to provide ‘intelligent input’ to our robotic platform, as a key contribution towards implementing fully automated organoid culture handling. Keywords: High-content imaging, Computational image analysis, Image recognition, Machine learning Supervisor: Gaspard Pardon Co-supervisor: Donati / Ando, EPFL Center for Imaging Contact: [email protected] Required: Computer programming (Python), Image analysis Availability: Open |
Computational Biology Interdisciplinary |
| Merten | Development of 2nd generation droplet microfluidic single-cell RT-PCR chips RT-PCR is a key step in single-cell analysis, including scRNAseq and paired sequencing of antibody encoding genes. However, its performance is limited by cellular inhibitory factors, which have a particularly high impact in low volume setting. The goal of the proposed master project is to overcome these limitations making use of advanced microfluidic and molecular biology approaches. During the project, the successful candidate will gain hands-on experience with microfluidic techniques and single-cell RT-PCR technology. In addition, the student will have the opportunity to become co-author and/or co-inventor on publications and patents resulting from the work. We are particularly looking for students with a background in molecular biology and a strong interest in interdisciplinary science. Prior expertise in microfluidics is a plus, but not absolutely required. Candidates are invited to send their applications by email to [email protected]. Keywords: microfluidics, single-cell analysis, RT-PCR Supervisor: Christoph Merten Contact: [email protected] Required: Molecular biology and a strong interest in interdisciplinary science Availability: Open |
Molecular biology Wet |
| Courtine | Identification of Brain Reorganizations in Advanced Stages of Parkinson’s Disease to Unlock New Neuromodulation Designs Neurorestore is a laboratory focusing on developing and applying medical therapies to restore neurological functions, integrating neuro-technologies with innovative treatments. Parkinson’s disease is currently incurable, and gait and balance impairments severely affect patients’ ambulation and independence with currently no clinical solution. To address this, the lab plans to use a transgenic mouse model of PD to identify new deep brain stimulation (DBS) targets that can alleviate these symptoms. The mouse model reproduces many aspects of progressive parkinsonism, including gait and balance impairments, allowing the lab to identify regions of interest with transcriptional activity and connectivity changes related to the emergence of these impairments. By surveying the entire brain and brainstem, the lab hopes to identify new DBS targets that can alleviate these symptoms. A Master’s student will have the opportunity to work with state-of-the-art techniques and models in neuroscience, whole-brain processing, behavioural task design, and analysis. The lab is seeking a person who can work on complex topics and assimilate multiscale information quickly. The Master’s student will help clarify the neuro-pathophysiology underlying gait and balance disorders in PD, leading to the development of effective DBS targets and treatments that will ultimately improve the lives of people with PD. Keywords: deep brain stimulation, parkinson’s disease, whole-brain, gait Supervisor: Gregoire Courtine Contact: [email protected] Required: immuno-histochemistry, data processing and analysis (Python, R and/or Matlab) Availability: Open |
Neuroscience Wet |
| Courtine | Transcutaneous Electrical Spinal Cord Stimulation in Rodents The profound transformation in the lives of people suffering from spinal cord injury well known. Our laboratory has been working with animal models to better understand how electrical spinal cord stimulation and rehabilitation can help patients recover sensorimotor function, with the ultimate goal of improving treatment and enhancing the quality of life of patients. Transcutaneous Electrical Stimulation (TES) has emerged as a non-invasive method for stimulating the spinal cord. Numerous studies have been conducted to comprehend how this stimulation aids individuals with spinal cord injuries (SCI) in recovering sensorimotor functions. However, thus far, there is no biological evidence to substantiate such improvements. In this role, you will support our research efforts to unveil the underlying mechanisms behind this type of stimulation. As a valued team member, you will have the opportunity to contribute by conducting extensive histological work. Your responsibilities will include utilizing state-of-the-art, high-resolution microscopes (such as light sheet microscopy and confocal microscopy) to analyze neuronal tissue. Additionally, you will have access to cutting-edge machinery for tissue slicing. To further enhance your analysis, you will employ fluorescent antibodies to stain the tissue, and employ computational tools to quantify the results. By joining our team, you will be at the forefront of scientific innovation, making significant contributions to the field. This is an exceptional opportunity to collaborate with leading experts, gain invaluable hands-on experience with advanced research techniques, and be part of a project that has the potential to transform lives. Keywords: spinal cord, rodent, stimulation Supervisor: Victor Perezpuchalt Contact: [email protected] Required: Histology experience, Microscopy, Image analysis, Programming experience in: Matlab, Python and or Java and R Availability: Open |
Neuroscience Dry and wet |
| Courtine | Neuronal axon regenera-on to restore upper-limb functions after paralysis Our team has managed to regenerate axons of propriospinal neurons across a complete spinal cord injury (SCI) for the first -me (Nature, MA Anderson et al. 2018). We currently focus on harnessing our regenerative therapy to recover neurological functions after SCI, including upper-limb movement, and to making it clinically applicable. The student will be involved in our effort to regenerate neuronal axons across a cervical SCI site to restore upper-limb functions after paralysis in rodents. Keywords: regeneration, paralysis, SCI, upper-limb Supervisor: Achilleas Laskaratos Contact: [email protected] Required: Preferred skills: previous research experience, cryostat tissue sectioning, immunohistological stainings, microscopy, image analysis, rodent perfusions, rodent animal work, R programing, biology and neuroscience knowledge Availability: Open |
Neuroscience Wet |
| Bernier-Latmani | Microdroplet-aided isolation of arsenic-methylating microorganisms from soil Arsenic is a toxic element that is widely distributed in the environment and mainly occurs in inorganic compounds. However, some soil microorganisms can transform these into organic (methylated) compounds. This is problematic in agricultural soils such as in rice paddy fields because the methylated arsenic can be taken up by the plant, affecting crop health and food safety. Although arsenic methylation is known to be particularly strong in flooded (i.e., oxygen-deprived) soils, very little is known about this process e.g., which anaerobic microorganisms (bacteria, archaea) have this unusual phenotype, what controls it (biochemical pathway), what benefit the cell gains from it, etc. This has remained elusive because it is difficult to isolate these specific organisms from their habitat, and thus they have never been studied in the lab. Thus, we seek to isolate arsenic-methylating microorganisms from a paddy soil. To do so, we recently developed a novel approach combining droplet-based microfluidics, arsenic whole-cell biosensors and FACS. The goal of this project is to optimize and apply this approach to isolate and further characterize new isolates. You will use diverse tools including microfluidics, molecular biology, microscopy and bioinformatics. Keywords: arsenic, biosensor, microdroplet, FACS Supervisor: Rizlan Bernier-latmani Contact: [email protected] Required: – Availability: Open |
Bioengineering Wet |
| La Manno | Developing a genomic deconvolution-based differential gene expression The field of biostatistics heavily relies on Differential gene expression analysis (DGEA) for deciphering the environmental effects and associated condition shifts. However, the conventional approach using pseudobulk data, randomly sampled from cells, has limitations in capturing donor-specific signals when pooling cells from multiple sources. In this master’s project, you will utilize the power of genotype-based clustering technology, introduced by souporcell, to develop a new approach, single-cell donor deconvolution differential gene expression (scDDDGE). This strategy aims to enhance statistical power through donor-guided pseudobulk aggregation. Your primary tasks will include determining the number of donors in a dataset, performing cell deconvolution, and identifying differentially expressed genes (DEGs). Preliminary research indicates scDDDGE may outperform traditional aggregation methods, providing improved DEG identification accuracy, even when cell numbers and donor representation are minimal. The primary goal of this master’s project is contribute developing and use scDDDGE to study inter-donor treatment response and identify robust cell type markers. This project offers the exciting potential to boost statistical power in experimental designs and contribute to our understanding of interindividual variation in development. Keywords: biostatistics, single-cell, computation, transcriptomics Supervisor: Gioele La Manno Contact: [email protected] Required: background in statistics and data science Availability: Open |
Computational Biology Dry |
| Herzog | Assessing the usefulness of prosthetic vision in a social virtual reality (VR) study We are currently looking for a student interested in conducting a virtual reality (VR) study in a master semester project. The study aims at assessing to what extent the POLYRETINA visual implant would allow future patients to master social scenarios. To this end, a social VR scenario is currently under development. Participants will be equipped with a VR headset and will see their virtual environment with artificial vision, i.e., as if they were a blind patient with the implant. With this vision, they will have to solve social tasks such as recognizing the positions of virtual avatars in the room. The project will consist in running a pilot study, if necessary, adapting the study protocol and, subsequently, in conducting the actual experiment. Experience with VR, Unity, and with conducting experiments on human participants is appreciated but not required. If you feel like making VR your research project sounds interesting, you will probably like it. : ) In case of interest, please email [email protected]. Happy to hear from you! Keywords: virtual reality, simulated prosthetic vision, artificial vision, experiment on sighted participants Supervisor: Sandrine Hinrichs Contact: [email protected] Required: – Availability: Open |
Neuroscience Dry |
| Dal Peraro | 3D density map embedding for cryoEM Cryo-Electron Microscopy (cryoEM) has opened unprecedented vistas in biology, offering detailed insights into macromolecular architectures. CryoEM captures snapshots of large complexes in their native configurations, enabling scientists to visualise their three-dimensional structures with unprecedented precision. However, the raw data is inherently noisy, and converting these 2D images into accurate 3D reconstructions demands meticulous postprocessing and advanced computational techniques. However, the challenge persists in dealing with various global and local resolutions and deriving precise structures from the cryoEM density maps. Owing to the inherently dynamic nature of proteins and their potential interactions with ligands and other subunits within complex systems, the task of effectively juxtaposing experimental cryoEM data with established reference structures presents an ongoing challenge. This project aims to construct density map embeddings encapsulating the information in the voxel grid through an autoencoder and evaluate their biological meaningfulness. We seek a motivated Master’s student, with a solid background in computer vision and deep learning, and a keen interest in addressing challenges in the realm of structural biology and protein modelling. The project is a collaboration between the Laboratory for Biomolecular Modeling (led by Prof. Dal Peraro) and the Computer Vision Laboratory (led by Prof. Fua). Keywords: computer vision, deep learning, cryoEM Supervisor: Matteo Dal Peraro Co-supervisor: Pascal Fua (IC) Contact: [email protected] Required: Data Science background (prior experience in Computer Vision would be prefered), with programming skills in Python, analytical skills and appetence for biology/bioengineering Availability: Open |
Computational Biology Interdisciplinary |
| Barth | AI-based generative approaches for de novo protein design AI-based approaches are revolutionizing structural biology and hold great promises for accelerating the discovery of therapeutics. However, despite tremendous advances, these methods have yet to generate de novo proteins for specific biomedical applications. To address these limitations, we are currently developing diffusion generative models to design de novo proteins with precise functions. We are seeking a masters student to work at the intersection of bioengineering and artificial intelligence. In this cutting-edge project, you will help develop and interrogate deep-learning models to design proteins with specific biomedical applications, with the ultimate goal of revolutionizing diagnostics and therapeutics. Your primary mission will be to help curate a comprehensive protein structural training dataset and assist in refining our existing deep learning models based on this dataset to generalize de novo protein design across the vast protein structure-sequence space, ultimately working towards solving one of synthetic biology’s greatest challenges. Working closely with the model’s author, you’ll receive one-on-one mentorship, and have the opportunity to improve your skills in biophysics, bioinformatics, AI and big data. Keywords: Protein design, deep learning, diffusion generative models, bioengineering, proteomics, personalized medicine Requirements: Proficiency in Python and bash is essential, while experience with protein structures and the PDB, along with knowledge of PyTorch and working on supercomputing clusters is highly relevant. Supervisor: Patrick Barth Contact: [email protected] Keywords: Bioengieering, Protein design, deep learning, diffusion generative models, bioengineering, proteomics, personalized medicine Supervisor: Patrick Barth Contact: [email protected] Required: Proficiency in Python and bash is essential, while experience with protein structures and the PDB, along with knowledge of PyTorch and working on supercomputing clusters is highly relevant. Availability: Open |
Computational Biology Dry |
| Altshuler | Molecular adaptation strategies of alpine snow microorganisms to warming This project will involve isolation of alpine snow microbial members, followed by genomic analysis to determine their metabolic potential in cold and warming adaptation. This will be complemented with cultivation if isolates at a gradient of temperatures and isolation of their RNA/proteins to determine the molecular changes triggered by changes in temperature. This work will be used to asses the thermal tolerances of cryospheric microorganisms. This project will also include an bioinformatic and analysis component, but will heavily rely on initial lab-work. Depending on the preference of the student, there is also an option to incorporate filed work into the project. Keywords: microbiology, gene expression, sequencing, culturing Supervisor: Ianina Altshuler Contact: [email protected] Required: General laboratory experience is a bonus, some microbiological backround Availability: Open |
Molecular biology Dry and wet |
| Altshuler | Composition and function of crysphere microbiomes This project will involve analysing in-house and publicly available metagenomes to determine their taxonomic compositions and functions and identify global genomic features of cryo-microbiomes. This would be followed up with a meta-annalysis to determine trends, patterns, and emergent qualities of microbiomes across different cryosphere environments. Keywords: microbiology, sequence data, metagenomics, genomics, microbiome Supervisor: Ianina Altshuler Contact: [email protected] Required: data analysis, working with large datasets, sequence analysis is a bonus Availability: Open |
Infectious diseases Dry |
| Dal Peraro | Graph neural network for prediction of protein-ligand binding affinity In the realm of drug discovery, one crucial factor often determines the success or failure of drug candidates – their interaction with targeted proteins. Accurately predicting, quantifying, and interpreting protein-ligand interactions (PLIs) is paramount in pre-clinical drug development. While various experimental methods exist for PLI quantification, they are labour-intensive, time-consuming, and sometimes fall short in sensitivity, especially when dealing with proteome-wide interactions. Consequently, there’s a growing shift towards modelling methodologies as a complement or even replacement for experimental screening procedures. In recent years, the utilization of deep learning techniques has made significant inroads into drug discovery. Geometric deep learning and graph neural networks (GNNs) have emerged as particularly promising tools for capturing intricate relationships and hidden patterns in structural data, leading to successful applications in PLI prediction. However, encoding proteins and ligands into meaningful numerical features is not a straightforward task and demands meticulous design to achieve generalization power over unforeseen cases. The research proposal aims to refine and enhance a cutting-edge GNN methodology to predict PLIs by innovating in the design of protein and ligand descriptors, incorporating latest advances in graph neural network architectures, and employing graph explainability and inference techniques. We seek a motivated Master’s student, with a background in structural biology, pharmacology or bioengineering, and a keen desire to explore and deploy deep learning approaches. The project will be hosted at the Laboratory for Biomolecular Modeling (led by Prof. Dal Peraro). Keywords: deep learning, graph neural network, chemoinformatics Supervisor: Matteo Dal Peraro Contact: [email protected] Required: Background in structural biology, pharmacology or bioengineering. Proficiency in Python is required, as well as analytical skills and appetence for AI and deep learning (prior experience in data science would be appreciated but not mandatory). Availability: Open |
Computational Biology Dry |
| Galland | Biochemistry for the binding of bacterial cells with gold nanoparticles for Surface-Enhanced Raman Spectroscopy (SERS) Bacteria enriched with gold nanoparticles emit very strong optical signals (Raman scattering) which allows for their detection and precise identification. Binding gold nanoparticles to bacteria (Figure 1) in a non-specific way is a relatively unexplored scientific avenue and requires the judicious selection of binding agents, as well as careful control of parameters such as pH, temperature and incubation times. The aim of the project will be to develop and optimize protocols for bacteria-gold adsorption focused on cost-efficiency, sustainability and ease-of-use. Evaluation of these protocols will be achieved through optical microscopy, Transmission Electron Microscopy (TEM) and SERS, among other techniques. Throughout the project, the student will work on chemically modifying gold nanoparticles to increase their affinity to bacteria while maintaining the stability of gold nanoparticles in aqueous solutions. This work will include a wide array of chemical work and techniques from gold nanoparticles synthesis, surface functionalization and ligand exchange, spectroscopy measurements (SERS & UV/Vis), Zeta potential measurements, TEM, bacterial culture and sample preparation, etc. Keywords: nanoparticles, bactria, biochemistry, spectroscopy Supervisor: Marwan Elchazli Contact: [email protected] Required: – Availability: Open |
Bioengineering Dry and wet |
| Galland | Machine learning model development for the classification of bacterial strains using their Raman chemical fingerprint Bacterial cells illuminated with a laser (for Raman spectroscopy) emit a characteristic “chemical fingerprint” which can be collected and analyzed with machine learning methods. Running this chemical fingerprint through machine learning algorithms allow us to determine the specific strain under observation. Available methods for this kind of application include PCA, LDA, Neural Networks, etc. The goal of the project will be to test different machine learning approaches, particularly neural networks-based techniques, to classify bacterial cells using their Raman scattering spectra. Throughout the project, the student will use pre-processing and machine learning methods in Python to work with large amounts of spectral data and classify them. They will be training and testing models and evaluating their effectiveness at classifying bacterial strains through a variety of metrics. Depending on the progress of the project, opportunities for publications might arise, which the student will be encouraged to participate in if they wish. Keywords: bacteria, classification, machine learning, spectroscopy Supervisor: Marwan Elchazli Contact: [email protected] Required: – Availability: Open |
Computational Biology Dry |
| Courtine | Internship / Master project Currently, externally applied spinal cord stimulation is being studied in order to understand its mechanisms. Clinical studies have shown that this technique is being effective in helping spinal cord injured patients to regain sensorimotor functions ( “Non-invasive spinal cord electrical stimulation for arm and hand function in chronic tetraplegia: a safety and efficacy trial” ). However, thus far, there is no biological evidence to explain such improvements. As part of our group, you will support our research efforts to unveil the underlying mechanisms behind this type of stimulation. We offer a rich and multidisciplinary project where your responsibilities will include performing histological work (e.g. tissue sectioning, immunostaining, in-situ hybridization). You will analyze these tissues using high-resolution microscopy and then analyze and quantify these images using computational tools, like QuPath or Fiji. You will also have the opportunity to become involved in electrophysiological studies assessing the muscular response in animal models due to different stimulation protocols. In parallel, depending on your interest and the availability of resources, you may also participate in numerical simulations, working on highly realistic biophysical animal models. Keywords: Non-Invasive spinal cord stimulation in rodent model Supervisor: Victor Perezpuchalt Contact: [email protected] Required: Previous histological experience, Microscopy experience (light sheet and/or confocal microscopes), Data analysis and Image processing, Signal processing, Programming experience in Python and Matlab (or other), Be comfortable to work with animals (rodents) Availability: Open |
Neuroscience Dry |
| Courtine | Master project : Robotic Platform for Upper Limb Experiments in Mice To assess, evaluate and quantify improvements in motor function, robotic platforms are essential. A robotic platform has been developed previously with incomplete aspects that need to be taken into account. The main goal of this project is to finish the robotic platform and to fine tune the system to allow proper training with rats and mice. In terms of practical work, the student will perform : – Software implementation and optimization to control motors, deal with the communication between all the components, show data in real time and interact with user input through a touchscreen and save this data for future analysis. – Electronics: assemble different electronic components and modify, repair or create some parts, if needed. – Mechanics: modify, repair or create some parts if needed. This project will challenge you in various aspects of device development and prototyping, including the design of mechanical and electronic parts, manufacturing and assembling them using 3D printers and other classical manufacturing processes, programming the control of the robot and the communication of the different robot parts. The project will conclude with animal experiments to validate the robot (in collaboration with the supervisor). Keywords: Robotic Platform for Upper Limb Experiments in Mice Supervisor: Victor Perezpuchalt Contact: [email protected] Required: – Programming skills in: Python, C/C++, Matlab and/or Java – Electromechanical prototyping experience – 3D design using: SolidWorks, Catia, Creo PTC (or similar) – 3D printing experience – Previous rodent animal work or be able to work with animals Availability: Open |
Bioengineering Dry |
Updated on August 15, 2024