CrysFormer: Protein Structure Prediction via 3d Patterson Maps and Partial Structure Attention

TL;DR

CrysFormer is a novel transformer-based model that leverages protein crystallography and partial structure data to predict protein electron density maps more efficiently than previous sequence-only methods.

Contribution

It introduces the first model to incorporate crystallography and partial structures into protein prediction, reducing data requirements and computational costs.

Findings

Achieves accurate predictions on peptide datasets

Uses smaller datasets than existing methods

Reduces computational costs significantly

Abstract

Determining the structure of a protein has been a decades-long open question. A protein's three-dimensional structure often poses nontrivial computation costs, when classical simulation algorithms are utilized. Advances in the transformer neural network architecture -- such as AlphaFold2 -- achieve significant improvements for this problem, by learning from a large dataset of sequence information and corresponding protein structures. Yet, such methods only focus on sequence information; other available prior knowledge, such as protein crystallography and partial structure of amino acids, could be potentially utilized. To the best of our knowledge, we propose the first transformer-based model that directly utilizes protein crystallography and partial structure information to predict the electron density maps of proteins. Via two new datasets of peptide fragments (2-residue and 15-residue) , we demonstrate our method, dubbed \texttt{CrysFormer}, can achieve accurate predictions, based on a much smaller dataset size and with reduced computation costs.