Machine learning in biological physics: From biomolecular prediction to design

Abstract

Machine learning has been proposed as an alternative to theoretical modeling when dealing with complex problems in biological physics. However, in this perspective, we argue that a more successful approach is a proper combination of these two methodologies. We discuss how ideas coming from physical modeling neuronal processing led to early formulations of computational neural networks, e.g., Hopfield networks. We then show how modern learning approaches like Potts models, Boltzmann machines, and the transformer architecture are related to each other, specifically, through a shared energy representation. We summarize recent efforts to establish these connections and provide examples on how each of these formulations integrating physical modeling and machine learning have been successful in tackling recent problems in biomolecular structure, dynamics, function, evolution, and design. Instances include protein structure prediction; improvement in computational complexity and accuracy of molecular dynamics simulations; better inference of the effects of mutations in proteins leading to improved evolutionary modeling and finally how machine learning is revolutionizing protein engineering and design. Going beyond naturally existing protein sequences, a connection to protein design is discussed where synthetic sequences are able to fold to naturally occurring motifs driven by a model rooted in physical principles. We show that this model is “learnable” and propose its future use in the generation of unique sequences that can fold into a target structure.

Machine learning meets physics: A two-way street

“…For our example of AlphaFold, the transformer idea appears to be absolutely essential. A brief guide to how transformers fit in with more general machine learning concepts dating all the way back to the Hopfield associative memory model ( 14 ) is discussed in the paper by Martin et al ( 15 ).…”

“…But this is not the only important source of data. As described in the article by Martin et al ( 15 ), much of the progress in protein folding arose from the recognition that comparing sequences of the same protein in different organisms could enable one to obtain important information regarding the contact map. The contact map is a matrix representation of the chance that residues located some distance apart along the backbone were likely in the folded structure to be nearby in three-dimensional space.…”