Blind prediction of protein B-factor and flexibility

TL;DR

This paper presents a novel machine learning approach combining graph theory to accurately predict protein B-factors and flexibility without prior data, outperforming traditional linear models.

Contribution

It introduces a new method using multiscale weighted colored graphs and global features for blind prediction of protein B-factors, advancing structural bioinformatics.

Findings

Predictions are more accurate than traditional linear fitting methods.

The approach is validated on hundreds of thousands of experimental B-factors.

Global features improve cross-protein B-factor prediction.

Abstract

Debye-Waller factor, a measure of X-ray attenuation, can be experimentally observed in protein X-ray crystallography. Previous theoretical models have made strong inroads in the analysis of B-factors by linearly fitting protein B-factors from experimental data. However, the blind prediction of B-factors for unknown proteins is an unsolved problem. This work integrates machine learning and advanced graph theory, namely, multiscale weighted colored graphs (MWCGs), to blindly predict B-factors of unknown proteins. MWCGs are local features that measure the intrinsic flexibility due to a protein structure. Global features that connect the B-factors of different proteins, e.g., the resolution of X-ray crystallography, are introduced to enable the cross-protein B-factor predictions. Several machine learning approaches, including ensemble methods and deep learning, are considered in the present work. The proposed method is validated with hundreds of thousands of experimental B-factors. Extensive numerical results indicate that the blind B-factor predictions obtained from the present method are more accurate than the least squares fittings using traditional methods.