GuideArtifEvol · Tracking and guiding artificial enzyme evolution via landscape inference

Directed evolution is an effective method for protein engineering. The host laboratory is designing innovative techniques for the directed evolution experiments of enzymes, with the possibility to test millions or billions of variants to be tested at each cycle. In this project, I will introduce the use of full-gene next generation sequencing to supervise the exploration of variants at the sequence level and quantitatively characterize the fitness landscape, with a focus on non-additive effects of mutations into fitness (epistasis). Furthermore, this information will be included in DE protocols to guide a more efficient exploration of the sequence space. The project is a strongly interdisciplinary project consisting of experimental synthetic biology, physics modelling and bioinformatics analysis. It is focused in the study of DNA processing enzymes such as polymerase and endonucleases. The proposal includes the development of specific tools which should have an immediate impact on the scientific community: i. extend the applications of next generation sequencers to characterize full length genetic libraries of protein variants ; ii. adapt quantitative methods, previously developed for sets of structurally homologous proteins and derived from Potts model to study the catalytic properties of during directed evolution. iii. develop a new experimental protocol to control the number of mutations in libraries. The analysis will provide a better understanding of the topology of the fitness landscape of an enzyme, and its distortions under selective pressures, thereby clarifying the relations between sequence,function and evolution in proteins.