Remarkably few vertebrates display tumor resistance properties. This select group includes salamanders, renowned for their tumor resistance, yet the molecular mechanisms underlying this ability remain unknown. Inspecting the attributes of salamanders reveals their massive genomes, ranging from 4.5 to 43 times larger than the human genome, as a unique biological feature. The sheer volume of genetic material in salamander cells makes the task of cell division complex, in terms of replication and also energetic demands. What are the advantages associated with possessing a giant genome? Our preliminary results suggest that giant genome size has imparted pressures that shaped novel innovations in cell cycle control. These innovations may be linked to the noted tumor resistance observed in newts.
The project: Others and we have created various scRNAseq datasets from animals with various genome sizes, different tissues, and across contexts (e.g., development versus adult homeostasis). The susceptibility to cancer and requirements for cell cycle regulation vary across these contexts. This project entails generating convolutional neural networks for coexpression to define the cell cycle regulome and its plasticity across cell types, lifestage, and phylogeny. The goal is to define a core, evolutionary conserved cell cycle regulome and subnodes of this regulome that are deployed across various species, genome sizes, and cell types. Moving forward, we will embed the neural-network-derived gene regulatory network into an agent-based model (ABM) aimed at quantitating the network resilience to specific gene perturbations.
The student: Master student (preferably 60cr thesis) with experience in bash, R or python. Experience in Java and/or neural networks and deep learning is advantageous.
What to expect: You will become well versed in quality control of publicly available data, working with various non-human species datasets, data integration, and workflow management. You will learn about cell cycle biology and explore its divergences across species. Depending on progress and interest you may also delve into generating neural networks and their integration in ABMs.
If this sounds of interest, please email nicholas.leigh@med.lu.se and virginia.turati@med.lu.se with a short motivation for what interests you about this project and your CV.