EDCB Open positions

https://www.epfl.ch/labs/barth-lab/

Expanding the universe of protein functions for synthetic biology and biomedicine

Our lab is developing and applying hybrid AI-based computational/experimental approaches for
engineering classes of proteins with novel functions for cell engineering, synthetic biology and
therapeutic applications. Through our bottom up design approach, we also strive to better
understand the molecular and physical principles that underlie the emergence, evolution and
robustness of the complex functions encoded by proteins and their associated networks.

We are part of RosettaCommons (https://rosettacommons.org/), a collaborative network of
academic laboratories that develop the software platform Rosetta and AI-based approaches for
macromolecular modeling and design. Ultimately, we aim to develop a versatile tool for designing
novel potent, selective therapeutic molecules, synthetic proteins, receptor biosensors, networks
and pathways for reprogramming cellular functions. We are also affiliated to the Ludwig Institute
for Cancer Research in Lausanne.

Projects in the lab are often multidisciplinary and involve the development of novel methods (e.g.
Feng, Nat Chem Biol 2016; Nat Chem Biol 2017; Paradis, Nat Comm 2022; Sengar, NeurIPS
2025) and their application involving experimental studies (e.g. Chen, Nat Chem Biol 2020; Yin,
Nature 2020; Jefferson, Nat Comm 2023; Chen, Nat Chemistry 2025). Projects involving external
collaborations with other research groups around the world or internal collaborations between
computational biologists, physicists and experimentalists in the lab are frequent. We also actively
translate our findings to the clinic in collaboration with physicians (Dr. Arber, Ludwig Institute for
Cancer Research, see Rath, Nat Biomed Eng 2025). Specific research topics include: 1. The
design of protein biosensors, mechanosensors and signaling receptors for reprogramming cell
(e.g. CAR T cell) functions and enhance cell-based therapies; 2. The design of highly selective
and potent protein and peptide-based therapeutics towards challenging targets such as GPCRs
or ion channels; 3. The study, prediction and design of protein dynamics and allostery using AI
and classic computational approaches; 4. The development of novel AI-based algorithms for
modeling & design of protein structures, interactions and motions.

Dry lab candidates should have strong programming skills in python/C/C++ and expertise in the
development of deep learning methods. Knowledge in structural biology, bioinformatics,
computational biomolecular modeling including molecular dynamics simulations is a plus.
Candidates more oriented towards the wet lab should have strong skills in molecular and cell
biology including experience in protein biochemistry, mammalian cell culture, microscopy, and
structural biology. Hybrid computational-experimental projects are also possible.

https://www.epfl.ch/labs/pbl/

PhD Position in Geometry-Guided Targeting at Biological Interfaces

The Programmable Biomaterials Laboratory (PBL) at EPFL is seeking a highly motivated PhD student in computational and quantitative biology to develop next-generation models and simulations for geometry-guided interactions at biological interfaces.

Project Vision

Biology is fundamentally spatial. Increasing evidence shows that cells do not simply respond to the presence of ligands or receptors, but also to their precise nanoscale organization, spacing, geometry, and collective presentation. Our laboratory develops programmable biomaterials and DNA-based nanotechnologies to uncover how geometry controls biological recognition, signaling, and selectivity.

This PhD project will focus on building predictive computational frameworks that connect molecular organization to biological function. Working in close synergy with an experimental team, the student will help establish a new quantitative framework for “multivalent engineering,” where geometry becomes a programmable biological input.

Research Objectives

The PhD candidate will contribute to several interconnected research directions:

• Model macroscopic multivalent interactions of proteins and biomolecular assemblies
  Develop theoretical and computational models that describe how geometry, valency, flexibility, and spatial organization shape collective binding and signaling behavior at biological interfaces.
• Discover geometric organization principles at cellular interfaces
  Use quantitative analysis and simulation to identify emergent spatial patterns in immune synapses, receptor assemblies, and membrane-associated signaling systems.
• Predict geometry-matched candidates for targeted drug delivery
  Build predictive frameworks for identifying optimal multivalent architectures capable of selectively targeting diseased cells based on receptor density and spatial organization.
• Expand the MEDUSA platform toward higher-order geometries
  Extend our existing MEDUSA framework toward more complex spatial architectures, enabling programmable multivalent pattern recognition across multiple geometric dimensions.
• Develop AI/ML-guided prediction of MEDUSA aptamers
  Integrate experimentally generated datasets with computational modeling and machine learning approaches to predict high-performing geometry-sensitive aptamer systems.

Candidate Profile

We are looking for candidates with strong quantitative and interdisciplinary interests. Applicants should have experience in one or more of the following areas:

• Computational biology
• Biophysics
• Quantitative biology
• Statistical physics
• Machine learning
• Molecular simulation
• Bioinformatics
• Applied mathematics or physics

Experience with programming (Python, MATLAB, or similar) and data analysis is expected. Prior experience with molecular simulations, reaction-diffusion systems, graph/network models, or AI/ML approaches is highly valued but not strictly required.

Research Environment

The student will join a highly interdisciplinary and collaborative environment spanning:

• DNA nanotechnology
• biomaterials
• immunoengineering
• multivalent targeting
• quantitative biointerface science

The project combines close interaction between computational and experimental researchers, with access to advanced imaging, nanofabrication, and high-throughput experimental platforms.

About the Lab

The Programmable Biomaterials Laboratory develops programmable biomolecular systems to understand and engineer biological interfaces across scales, with applications in targeted therapeutics, immune engineering, and synthetic biology.

Location

The position is based at EPFL in Lausanne, Switzerland.

I actually have 2 different projects:

A Foundation Model and Knowledge System for Neurodevelopmental Biology Research

This PhD project addresses the growing challenge in neurodevelopmental biology where comprehensive molecular atlases remain largely inaccessible to researchers for routine experimental interpretation and hypothesis generation. Despite containing unprecedented cellular and molecular information, current atlases function as static databases that require specialized expertise to query effectively, limiting their practical utility for everyday laboratory research.

The project will develop a transformative AI system combining two innovative components: a cell-type-informed 3D foundation model that integrates spatial distributions with molecular profiles using dual-stream neural architecture, and NeuroBioKG, a literature-powered knowledge system that will mine decades of neurodevelopmental research to construct comprehensive knowledge graphs. The foundation model will be trained on millions of synthetic examples to enable complex tasks including gene expression prediction and experimental data augmentation through natural language queries, while the knowledge system will use graph retrieval augmented generation to provide contextual interpretation against established developmental principles.

Together, these components will transform molecular atlases from static resources into interactive research tools that actively support hypothesis generation and experimental design. The project is ideal for students with computational biology backgrounds within combining machine learning expertise with neurodevelopmental biology to create next-generation research AIs reaching expert-level insights.

Investigating Lipid Dysregulation in Neural Tube Defects and Metabolic Intervention Strategies

This PhD project addresses a critical gap in understanding how environmental teratogens disrupt neural tube development through lipid metabolism perturbations, and explores whether metabolic interventions can prevent or alleviate neural tube defects (NTDs). Despite growing evidence that teratogenic compounds affect lipid homeostasis during early brain development, the specific mechanisms linking lipid dysregulation to NTDs remain poorly characterized, limiting our ability to develop protective strategies.

The project will systematically investigate how teratogenic compounds induce lipid metabolic disruptions using state-of-the-art ex utero mouse embryo culture systems combined with spatial lipidomics and transcriptomics. You will expose developing embryos to well-characterized teratogens (including retinoic acid, methotrexate, and cyclopamine) and endocrine-disrupting chemicals, then use MALDI mass spectrometry imaging and spatial gene expression analysis to map region-specific changes in lipid composition and neural patterning. A key innovation will be testing whether metabolic modulators can rescue teratogen-induced defects by rebalancing disrupted lipid pathways.

The work will combine cutting-edge spatial multi-omics technologies with advanced computational modeling to identify metabolic intervention targets. Using machine learning approaches and biochemically-constrained models, you will predict which metabolic pathways could be therapeutically targeted to counteract specific teratogenic effects.

This project is ideal for students with backgrounds in developmental biology or biochemistry who are interested in combining experimental embryology with computational approaches to address clinically-relevant questions about birth defects and environmental health.

https://www.epfl.ch/labs/naef-lab/

Circadian Clock Synchrony in Health and Disease across Cells and Tissues

Life on Earth at all scales (societies, behavior, physiology, molecular functions) is temporally organized along the 24h daily cycle. Modern lifestyles in a 24/7 society disrupt this temporal organization, causing misalignment of body clocks with environmental cycles, increasing risk of cancer, metabolic and cardiovascular disease.

This project builds on our longstanding interest to understand circadian synchrony and gene regulatory mechanisms underlying circadian rhythms, and notably its impact on temporal physiology and disease states in tissues. We are looking for an interdisciplinary profile (proportions of wet and dry work to be defined and discussed with the candidate) to work on developing assays to define oscillator states and circadian synchrony from single cell measurements. Applications are in the areas of circadian cell communication, metabolic diseases and cancer.

https://www.epfl.ch/labs/ramdya-lab/

In the Neuroengineering Laboratory, we are reverse-engineering cognitive and motor behaviors in the fly, Drosophila melanogaster, to better understand the mind and to design more intelligent robots. Flies are an ideal model: they generate complex behaviors, their nervous systems are small, and they are genetically malleable. Our lab develops and leverages advanced microscopy, machine learning, genetics, and computational modeling approaches to address systems-level questions.

We are always looking for talented researchers to join our team. Join us! There is much to discover!”

https://www.epfl.ch/labs/upzenk/

PhD Position – Computational Epigenomics of Brain Development and Disease

We are looking for a highly motivated PhD student with a strong background in computational biology, bioinformatics, quantitative biology, or related fields to develop computational models of human brain development and neurodevelopmental disease.

The project will integrate large-scale single-cell transcriptomic and epigenomic datasets from the developing human brain and human stem cell-derived brain models to reconstruct developmental trajectories and identify the regulatory and epigenetic changes that accompany cell fate transitions.

A major goal will be to develop integrative computational models that connect developmental dynamics with changes in the epigenome, and to use these models to understand how normal developmental trajectories are altered in neurodevelopmental disorders.

https://piyushnanda-sysbio.github.io/

Investigating learning in signal transduction networks 

Outside the laboratory environment, abundance of glucose is an exception rather than the rule. Cells ferment in the presence of excess glucose (Warburg effect in cancer cells or Crabtree effect in yeast) and activate respiration when glucose is low. The switch is costly and slow, because activating respiration requires building a mitochondrial network. This raises the question whether glucose-sensing signal transduction networks have evolved to “learn” the history of glucose availability in the environment to strategically switch between respiration and fermentation.

The project stems from my Ph.D. research work (Nanda and Murray, 2026, bioRxiv) in which we used single-cell metabolic imaging of Saccharomyces cerevisiae across environments. We found that the glucose-sensing network Ras/cAMP/PKA alters the fraction of cells which are fermenting or respiring at a fixed glucose concentration. We will use natural isolates of S. cerevisiae collected from around the world. Each has evolved in a different ecological niche, so their genomes may encode different environmental histories.

In this project, we will transplant highly variable alleles (wild alleles) of the Ras/cAMP/PKA pathway from these isolates into a reporter strain that measures flux through glycolysis and mitochondrial activity in individual cells. We will assay how these wild alleles in signal transduction circuits govern the switch between fermentation and respiration during fluctuations from high glucose to low glucose. We will also develop a high-throughput barcoded pooled assay that measures the fitness of the strains carrying these alleles in custom time-series of glucose fluctuations in an automated bioreactor. We will make single-cell measurements of metabolism, growth rate and lag phase in microfluidic chips to track cells during this transition for selected combinations of the wild alleles.   

Together, we will use standing natural variation in a signaling circuit to find out whether cells encode their environmental history to promote rapid adaptation to change and uncover what principles govern such learning.

Candidates with experience in microbiology and molecular biology are encouraged to apply. Familiarity with Python programming and an interest in computational biology is a plus.

Any questions:

[email protected]