Learning probability distributions from data: applications to Protein Design and predictions of Mutational effects.
In this talk I will introduce the inference of a graphical model, also called Potts model in statistical physics, from sequence data of a protein family [1].Such models help in establishing the mapping between the sequence of a protein and its structure and function.
I will present two applications of such modelling, related to recent and ongoing works. The first is on protein design, in which the graphical model build on sequence data of natural proteins is used to engineer new artificial protein which functionality has been experimentally tested by our collaborators [2]. The second is the prediction of single site mutational effect around a wild-type sequence, which is tested on single mutational scan data of several wild-types in different protein families.
[1]Inverse Statistical Physics of Protein Sequences: A Key Issues Review
S. Cocco, C. Feinauer, M. Figliuzzi, R. Monasson, M. Weigt
Reports on Progress in Physics 81, 032601 (2018).
[2] An evolution-based model for designing chorismate mutase enzymes
W.P. Russ, M. Figliuzzi, C. Stocker, P. Barrat-Charlaix, M. Socolich, P. Kast, D. Hilvert, R. Monasson, S. Cocco, M. Weigt, R. Ranganathan .Science 369, 440-5 (2020).Connect to the seminar:https://us02web.zoom.us/j/83880710317?pwd=dVA5NHB6UXlTM3NsVWh2bzZTMHhEZz09
ID de réunion : 838 8071 0317
Code secret : 692394
- This évènement has passed.
