Molecular Docking: Methodological Approaches of Risk Assessment

Abstract

Introduction. Computational chemistry methods and, particularly, the noncovalent molecular docking are increasingly implemented into the practice of drug development. Previously, a risk management of potential biases did not applied for this relatively young research instrument. Aim. The study objective was to design the risk assessment system for noncovalent molecular docking. Materials and methods. The development of bias risk assessment system was based on the world's leading practices in noncovalent molecular docking. Results and discussions. As a result of the deductive analysis of the molecular docking process, bias domains were identified and a risk-based algorithm was proposed, which was tested on a sample of articles obtained during a systematic review. A tendency to frequent limited provision of information on the methodology of the computational experiment, as well as on the application of practices proven to lead to irrelevant results of molecular docking, has been revealed. Conclusion. The data obtained cannot be extrapolated to all studies that refer to the results of molecular modeling. However, through the proposed risk-based algorithm, the attention of researchers is focused on assessing the quality of such publications. We hope that the developed tool for bias risk assessment in noncovalent molecular docking will be finalized and eventually put into practice. It will possibly reduce the share of low-quality work in the field of drug development at the earliest stages.