PrankWeb 4: a modular web server for protein-ligand binding site prediction and downstream analysis.

Journal: Nucleic acids research
Published Date:

Abstract

Knowledge of protein-ligand binding sites (LBSs) is crucial for advancing our understanding of biology and developing practical applications in fields such as medicine or biotechnology. PrankWeb is a web server that allows users to predict LBSs from a given three-dimensional structure. It provides access to P2Rank, a state-of-the-art machine learning tool for binding site prediction. Here, we present a new version of PrankWeb enabling the development of both client- and server-side modules acting as postprocessing tasks on the predicted pockets. Furthermore, each module can be associated with a visualization module that acts on the results provided by both client- and server-side modules. This newly developed system was utilized to implement the ability to dock user-provided molecules into the predicted pockets using AutoDock Vina (server-side module) and to interactively visualize the predicted poses (visualization module). In addition to introducing a modular architecture, we revamped PrankWeb's interface to better support the modules and enhance user interaction between the 1D and 3D viewers. We introduced a new, faster P2Rank backend or user-friendly exports, including ChimeraX visualization.

Authors

  • Lukáš Polák
    Department of Software Engineering, Faculty of Mathematics and Physics, Charles University, Prague, 121 16, Czech Republic.
  • Petr Škoda
    Department of Software Engineering, Faculty of Mathematics and Physics, Charles University, Ke Karlovu 3, Prague 2, Czech Republic.
  • Kamila Riedlová
    Department of Software Engineering, Faculty of Mathematics and Physics, Charles University, Prague, 121 16, Czech Republic.
  • Radoslav Krivak
    Department of Software Engineering, Faculty of Mathematics and Physics, Charles University, Czech Republic.
  • Marian Novotny
    Department of Cell Biology, Faculty of Science, Charles University, Czech Republic.
  • David Hoksza
    Department of Software Engineering, Faculty of Mathematics and Physics, Charles University, Ke Karlovu 3, Prague 2, Czech Republic.