Predicting the real-valued distances between residue pairs for proteins

TL;DR

This paper introduces a regression-based method using generative adversarial networks to predict continuous real-valued residue-residue distances in proteins, improving structure modeling accuracy and applicability to membrane proteins.

Contribution

It presents a novel regression approach with GANs for residue distance prediction, surpassing classification-based methods and enabling direct membrane protein structure prediction.

Findings

Achieves comparable or better accuracy than state-of-the-art methods on CASP13 targets.

Enables rapid structure modeling using predicted distances with CNS.

Effective for membrane proteins without transfer learning.

Abstract

Predicting protein structure from the amino acid sequence has been a challenge with theoretical and practical significance in biophysics. Despite the recent progresses elicited by improved residue-residue contact prediction, contact-based structure prediction has gradually reached the performance ceiling. New methods have been proposed to predict the residue-residue distance, but unanimously by simplifying the real-valued distance prediction into a multiclass classification problem. Here we show a regression-based distance prediction method, which adopts the generative adversarial network to capture the delicate geometric relationship between residue pairs and thus could predict the continuous, real-valued residue-residue distance satisfactorily. The predicted residue distance map allows rapid structure modeling by the CNS suite, and the constructed models approach at least the same level of quality as the other state-of-the-art protein structure prediction methods when tested on available CASP13 targets. Moreover, this method can be used directly for the structure prediction of membrane proteins without transfer learning.