Abstract

Proteins are versatile molecules that can display a wide range of behaviours, especially in vivo, where for example post-translational modifications can significantly affect their shape, interactions, and function. Such behaviour is not readily captured by their sequence, nor by structural biology and/or molecular dynamics approaches. It instead necessitates a more probabilistic perspective of proteins that describes their potential ‘emergent' behaviour, rather than context-dependent behaviour (e.g. a final folded state). The framework we developed aims to capture such ‘biophysical features’ of proteins, such as backbone and side-chain dynamics (DynaMine), conformational propensities, early folding (EFoldMine), disorder (DisoMine), and beta-sheet aggregation (Agmata) (https://bio2byte.be/b2btools/). Through a set of examples, we illustrate how this framework is being used to define ‘biophysical conservation’ in evolution, as well as interpreting protein behavior such as secretability or early folding.

(pass code:^=PemP3P)