Assessment of accuracy of AlphaFold protein models for ligand docking
Embargo Date
2027-01-28
OA Version
Citation
Abstract
Molecular docking is a powerful computational tool for predicting protein-ligand interactions, widely employed in drug discovery. However, its effectiveness is often constrained by the availability of experimentally resolved X-ray protein structures, a process that is both time consuming and resource-intensive. AlphaFold2 (AF), a machine learning (ML)-based method, offers an efficient alternative by predicting high-accuracy 3D protein structures directly from amino acid sequences. This study assesses the utility of AF-generated protein models for fragment and larger ligand docking with Glide, a widely used docking approach developed by Schrödinger Inc. The docking workflow is evaluated in an unbiased manner by leveraging binding site identification with FTMap, a binding hot spot prediction software. Glide flexibly docks ligands to a rigid protein structure which can hinder its accuracy by the lack of protein flexibility–an essential factor for ligand binding. This thesis work systematically explored strategies for incorporating protein flexibility to improve the docking outcomes. The results will provide insights into the effectiveness of integrating AF protein models into docking procedures, highlighting the potential for streamlining computational drug discovery processes.
Description
2025