Overview:
Our lab is interested in exploring the interactions between the cell cycle and the circadian clock, two fundamental cyclic processes in cellular biology (Bieler et al., MSB, 2014; Droin et al., Nature physics, 2019). We have engineered mammalian cells containing fluorescent reporters for both the cell-cycle and the circadian-clock, allowing real-time observation of these processes through live microscopy imaging. However, inferring phases solely based on these reporter signals has proven to be intricate. Therefore, we believe that analyzing the entire images to extract additional cellular features could provide a richer understanding and accurate phase identification.
Aims:
The core aim of this project is to construct an optimal latent space for cell cycle and the circadian clock states using the live imaging data. The student will develop and apply autoencoder models to process the live imaging data, with the goal of accurately identifying the underlying phases of the cell-cycle and the circadian clock. Moreover, the student will investigate specific cellular features such as changes in nuclei shape or cell membrane, breakdown events, chromatin compaction to understand how they contribute to the identified latent spaces and if they can aid in predicting the phases.
Prerequisites:
- A solid understanding of deep learning and image analysis is needed.
- Basic knowledge of cell biology is helpful but not mandatory.
- Familiarity with live microscopy imaging and image processing techniques would be an added advantage.
Contact: [email protected]