Open Drug Discovery Toolkit (ODDT): a new open-source player in the drug discovery field

Maciej Wójcikowski,Pawel Siedlecki,Piotr Zielenkiewicz

doi:10.1186/s13321-015-0078-2

Abstract

BackgroundThere has been huge progress in the open cheminformatics field in both methods and software development. Unfortunately, there has been little effort to unite those methods and software into one package. We here describe the Open Drug Discovery Toolkit (ODDT), which aims to fulfill the need for comprehensive and open source drug discovery software.ResultsThe Open Drug Discovery Toolkit was developed as a free and open source tool for both computer aided drug discovery (CADD) developers and researchers. ODDT reimplements many state-of-the-art methods, such as machine learning scoring functions (RF-Score and NNScore) and wraps other external software to ease the process of developing CADD pipelines. ODDT is an out-of-the-box solution designed to be easily customizable and extensible. Therefore, users are strongly encouraged to extend it and develop new methods. We here present three use cases for ODDT in common tasks in computer-aided drug discovery.ConclusionOpen Drug Discovery Toolkit is released on a permissive 3-clause BSD license for both academic and industrial use. ODDT’s source code, additional examples and documentation are available on GitHub (https://github.com/oddt/oddt).Electronic supplementary materialThe online version of this article (doi:10.1186/s13321-015-0078-2) contains supplementary material, which is available to authorized users.

Highlights

There has been huge progress in the open cheminformatics field in both methods and software development
The current version of our toolkit provides machine learning models that are widely used in cheminformatics and drug discovery: (1) random forests, (2) support vector machines, and (3) artificial neural networks
Example 1: filtering, docking and re‐scoring workflow In this code example, the researcher uses Open Drug Discovery Toolkit (ODDT) to dock a database of ligands with Autodock Vina and rescore the results with two independent scoring functions

Summary

Introduction

There has been huge progress in the open cheminformatics field in both methods and software development. Many novel computational chemistry methods were developed to aid researchers in discovering promising drug candidates. Much progress has been made in areas such as scoring functions, similarity search methods and statistical approaches (for review see [1, 2]). Some of the most popular and successful methods in drug discovery are structure-based. Structure-based methods are commonly employed to screen large smallmolecule datasets, such as online databanks or smaller sets such as tailored combinatorial chemistry libraries. These techniques, from molecular docking to molecular mechanics to ensemble docking, employ scoring processes that are crucial for decision making. Empirical scoring functions use explicit equations based on physical properties of available ligand-receptor complexes

Objectives

Methods

Results

Conclusion