FLINDIP · Fitness landscape of intrinsically disordered proteins

Fitness landscape of intrinsically disordered proteins

What is the problem/issue being addressed? In this project, we are focusing on intrinsically disordered proteins, a large group of proteins whose mechanism of action and role is poorly understood. Such proteins do not adopt a unique native structure. Instead, they explore numerous conformations depending on external conditions. Recent bioinformatic analyses show that up to 15% of all proteins are intrinsically disordered. result Why is it important for society? The critical role of intrinsically disordered proteins in cellular functions and in the onset of pathological conditions generated significant interest for their study. Much effort has been devoted to map the effects of particular mutations on protein functionality. However, there has been no attempt to systematically study the genotype-to-phenotype link in intrinsically disordered proteins. Obtaining such information is essential for improving theoretical models of protein folding and molecular evolution, as well as for de-novo design of intrinsically disordered proteins with improved activities. What are the overall objectives? The aim of this project is to experimentally measure and analyse, the genotype-to-phenotype connection for several intrinsically disordered proteins by deep mutational scanning. We are focusing on disordered proteins that contribute to surviving complete desiccation in their host organisms. The approach we follow is to generate libraries of cells expressing hundreds of thousands of variants of these proteins and to perform competition assays to measure functionality of every variant in the library. We then use a variety of approaches to analyse the dataset, including machine-learning algorithms to model the fitness of variants and to design new functional intrinsically disordered proteins. Conclusion of the action: The project was made of two work packages. The first package consisted in producing libraries of mutants of the target genes. Despite our best efforts the phenotype of the target genes previously published by another lab could never be reproduced so we decided to work on different genes. The our lab closed for a long period of time due to Covid-19. Upon resuming lab work, preliminary data on the new genes showed promising results. Production of new libraries is ongoing. The second work package is data analysis. Due to the amount of data produced by our approach we decided to tackle data analysis by using machine learning. Our collaborators at the Kondrashov lab shared a dataset similar to the one we plan to obtain (mutations versus phenotype).The results of this study were published, and the pipeline developed for this data is fully reusable on disordered proteins.

Data: CORDIS, © European Union