TL;DR
The interface prediction program WHISCY is presented, which combines surface conservation and structural information to predict protein–protein interfaces and demonstrates the potential of using interface predictions to drive protein–protein docking.
Blind docking is a method for predicting a binding mode of a ligand with a protein without any prior information about a binding site. Some tools allow this type of docking experiment directly, others, including some established tools, require binding site information being passed as an input. In this latter case, one can use cavity prediction tools and use the results of their prediction as an input in these docking calculations. However, it is still unclear if the results of these predictions can be reliably used in protein–ligand docking and what is the best technical way to pass this information to the docking algorithm. In this study we estimated the applicability of the binding pocket prediction tools in docking experiments to address this gap in knowledge. We use four different computational tools for cavity prediction and use the best predicted cavities represented in different ways to run GOLD docking calculations. Analysis of subsequent use in docking highlights that Fpocket and CAVIAR are the best performing cavity prediction tools in this context. Further analysis shows that accurate binding site input does not guarantee accurate binding pose predictions and, even with the predicted cavities, the more restrained the input is, the more reliable the docking results are.
A program with the DockAnalyzer graphical interface has been developed to automate the analysis of molecular docking results in CIF format. The need for such a tool is caused by the exponential growth of data during virtual screening, when manual processing of a large number of files becomes impossible, and existing solutions require complex configuration or transfer of confidential structures to external servers. The aim of the work is to create a lightweight desktop application for local analysis and visualization of intermolecular interactions in protein-ligand complexes. The program is implemented in Python using the GEMMI, NumPy, and Tkinter libraries. The architecture includes modules for parsing CIF files, geometric classification of contacts (hydrogen bonds, hydrophobic interactions, flares) and an interactive table with color coding of bond types. Additionally, integration with the P2Rank tool for predicting ligand-binding pockets is implemented, which allows comparing calculated contacts with predicted binding regions. Testing on the reference complex of HIV-1 protease with an inhibitor confirmed the correctness of the operation: 470 contacts were automatically identified, the distribution by type of interactions and the remnants of the binding site coincided with those annotated in the PDB. A comparative analysis with analogues (PLIP, Arpeggio, BINANA 2) showed that the advantages of DockAnalyzer are visual visualization of data directly in the contact table, the absence of dependencies on external servers, and confidentiality of processing due to local startup. The program exports the results to TXT and CSV formats for further statistical processing. The modular architecture makes expanded functionality possible, including support for new file formats and machine learning methods for predicting binding pockets.
COACH-D 2.0 is introduced, a substantially enhanced template-based method for predicting protein-ligand binding sites and features three key advances: integration of multimeric templates from Q-BioLiP into the authors' in-house library, a new multimeric structure processing module enabling binding site prediction for protein complexes, and an efficient template screening strategy that significantly boosts both prediction speed and accuracy.