Modeling sequence-space exploration and emergence of epistatic signals in protein evolution

TL;DR

This paper introduces stochastic models of protein evolution that predict features of experimental sequence libraries, enabling analysis and optimization of evolutionary experiments to detect epistatic signals and infer protein structure.

Contribution

The authors develop data-driven fitness landscape models that simulate protein evolution, providing a quantitative framework to analyze epistasis and optimize experimental design.

Findings

Models accurately predict fitness distributions and mutational spectra.

Large, diverged libraries are necessary to detect epistatic signals.

Framework can forecast experimental outcomes and guide protocol optimization.

Abstract

During their evolution, proteins explore sequence space via an interplay between random mutations and phenotypic selection. Here we build upon recent progress in reconstructing data-driven fitness landscapes for families of homologous proteins, to propose stochastic models of experimental protein evolution. These models predict quantitatively important features of experimentally evolved sequence libraries, like fitness distributions and position-specific mutational spectra. They also allow us to efficiently simulate sequence libraries for a vast array of combinations of experimental parameters like sequence divergence, selection strength and library size. We showcase the potential of the approach in re-analyzing two recent experiments to determine protein structure from signals of epistasis emerging in experimental sequence libraries. To be detectable, these signals require sufficiently large and sufficiently diverged libraries. Our modeling framework offers a quantitative explanation for the variable success of recently published experiments. Furthermore, we can forecast the outcome of time- and resource-intensive evolution experiments, opening thereby a way to computationally optimize experimental protocols.