Flexible protein–protein docking with a multitrack iterative transformer

Abstract

Conventional protein‐protein docking algorithms usually rely on heavy candidate sampling and re‐ranking, but these steps are time‐consuming and hinder applications that require high‐throughput complex structure prediction, e.g., structure‐based virtual screening. Existing deep learning methods for protein‐protein docking, despite being much faster, suffer from low docking success rates. In addition, they simplify the problem to assume no conformational changes within any protein upon binding (rigid docking). This assumption precludes applications when binding‐induced conformational changes play a role, such as allosteric inhibition or docking from uncertain unbound model structures. To address these limitations, we present GeoDock, a multi‐track iterative transformer network to predict a docked structure from separate docking partners. Unlike deep learning models for protein structure prediction that input multiple sequence alignments (MSAs), GeoDock inputs just the sequences and structures of the docking partners, which suits the tasks when the individual structures are given. GeoDock is flexible at the protein residue level, allowing the prediction of conformational changes upon binding. On the DIPS test set, GeoDock achieves a 43% top‐1 success rate, outperforming all other tested methods. However, in the standard DIPS train/test splits, we discovered contamination of close homologs in the training set. After decontaminating the training set, the success rate is 31%. On the DB5.5 test set and a benchmark dataset of antibody‐antigen complexes, GeoDock outperforms the deep learning models trained using the same dataset but falls behind most of the conventional methods and AlphaFold‐Multimer. GeoDock attains an average inference speed of under one second on a single GPU, enabling its application in large‐scale structure screening. Although binding‐induced conformational changes are still a challenge owing to limited training and evaluation data, our architecture sets up the foundation to capture this backbone flexibility. Code and a demonstration Jupyter notebook are available at https://github.com/Graylab/GeoDock.This article is protected by copyright. All rights reserved.

A Reversible Spherical Geometric Conversion of Protein Backbone Structure Coordinate Matrix to Three Independent Vectors of ρ, θ and φ

“…This feature is crucial for maintaining prediction accuracy and enabling further analysis of the extracted structural features for its future application in protein structure prediction [13][14][15][16][17][18]101]. 2. the compact representation offered by spherical coordinates with Caenopore-5 as an example enhances computational efficiency, making our method suitable for large-scale (for both PDB and AlphaFoldDB) protein structure prediction tasks for the entire protein molecular space [47,[102][103][104][105][106][107][108][109]. 3. lastly, the intuitive nature of spherical coordinates also aids in the interpretation of structural features, potentially offering insights into the underlying principles governing protein folding and function [16,[110][111][112][113][114][115][116][117][118].…”

Quantum Chemistry-Based Protein–Protein Docking without Empirical Parameters

“…Specifically, within the realm of protein science, significant efforts have been devoted toward the development of protein–protein docking methods − to predict the three-dimensional structures of protein complexes. Consequently, many efficient methods, − including easily accessible web tools, have emerged. This has also led to the development of benchmark sets for protein–protein docking, − alongside the establishment of the critical assessment of predicted interactions (so-called CAPRI) experiment, − which is a blind test of protein–protein docking.…”

“…Protein–protein docking methods can be broadly classified into two types: template-based docking − and ab initio docking. − Among these, the former leverages known structural information on highly homologous protein complexes, while the latter solely relies on the structural information on the given two proteins. Ab initio docking is further classified into flexible docking, − which considers the internal degrees of freedom of molecules, and rigid-body docking, − which does not. Flexible docking is likely preferable to rigid-body docking because proteins are generally flexible.…”