Virtual screening has been widely adopted as a valid and useful tool for reducing the cost, time and effort needed to identify candidate compounds for drug discovery. However, while there are free software tools available (i.e. AutoDock Vina) to perform virtual screening many of the currently available tools require the investment of extensive time and effort for the non-expert user. In this project we hypothesized that the virtual screening process could be streamlined by using scripts and freely available software to make these valuable tools easily accessible to non-expert users. Workflows for both Windows and Unix (e.g. MacOS and Linux) systems were developed, using Powershell and BASH scripts, respectively. These scripts were constructed to enable a non-expert user to generate a virtual compound library from PubChem, perform molecular docking using AutoDock Vina, re-score the docking results with DSX<sup>2</sup> and sort and store the results in a spreadsheet. These scripting tools are especially valuable for users who may lack the ability or time to develop their own virtual screening workflow, but more experienced users can also utilize these scripting tools by modifying them to suit their own needs. We estimated that these scripts reduce ~100 hours of tedious data input to a few minutes of data input followed by 2-3 hours of computational time for a database of 1700 compounds, which represents an ~600% reduction in the time needed to perform this virtual screening experiment.

Making virtual screening accessible to non-experts through automation of library generation, ligand screening, and binding pose scoring in a single workflow

2019 Spring National Meeting

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.