Characterizing Drug-Target Residence Time with Metadynamics: How To Achieve Dissociation Rate Efficiently without Losing Accuracy against Time-Consuming Approaches

J Chem Inf Model. 2017 Aug 28;57(8):1895-1906. doi: 10.1021/acs.jcim.7b00075. Epub 2017 Aug 11.

Abstract

Drug-target residence time plays a vital role in drug efficacy. However, there is still no effective strategy to predict drug residence time. Here, we propose to use the optimized (or minimized) structures derived from holo-state proteins to calculate drug residence time, which could give a comparable or even better prediction accuracy compared with those calculated utilizing a large number of molecular dynamics (MD) structures based on the Poisson process. Besides, in addition to the Poisson process, one may use fewer samples for predicting residence time due to the reason that, in a large extent, the calculated drug residence time is stable and independent of the number of samples used for the prediction. With remarkably reduced computational load, the proposed strategy may be promising for large-scale drug residence time prediction, such as post-processing in virtual screening (VS) and lead compound optimization.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Algorithms
  • Drug Discovery*
  • Molecular Dynamics Simulation*
  • Molecular Targeted Therapy*
  • Poisson Distribution
  • Protein Conformation
  • Proteins / chemistry
  • Proteins / metabolism*
  • Time Factors

Substances

  • Proteins