Antidote: An open platform for data-driven drug discovery applications

drawing

George Papadatos, Michal Nowotka, Nathan Brown, Greg Landrum and Patrick Walters

Executive Summary

Vision: we propose Antidote, an open source, state-of-the-art, end-to-end platform which enables the complete life cycle of compound data acquisition, curation, storage, querying, mining, dissemination and reuse for chemoinformatics and drug discovery applications. Specifically, Antidote combines open chemical, pharmacological and chemogenomics data and analogous in-house data with open source chemoinformatics and data science and mining tools, along with interactive demos, tutorials, documentation and drug discovery use-cases in a free, accessible and intuitive platform. We envisage that this will catalyse and facilitate

  1. Academic drug discovery via easy access to standardised infrastructure, data and use-cases;
  2. Accessibility and dissemination of modern informatics and computational applications for data-driven drug discovery to all levels of the scientific community;
  3. Interactive teaching and learning of chemoinformatics and data mining concepts and applications;
  4. Reproducibility and sharing of research methodologies and outcomes; and
  5. Synergy and collaborations between different institutions and domain experts outside drug discovery.

Implementation: As an existing early-stage prototype, we use myChEMBL, which was co-developed by members of our team at the EBI. myChEMBL is a virtual machine which hosts the ChEMBL database and a number of data analysis tools. We plan to heavily extend and scale-up myChEMBL by combining existing open data content, open source data mining and analytics technologies, as well as user-friendly data curation and querying web platforms and applications built by members of our team. Importantly, the resulting platform will be modular, scalable and accessible via several deployment methods.

Audience: We envisage that Antidote will become a widely adopted data analysis and reuse platform among researchers of diverse backgrounds and experience, ranging from students to experts and from wet-lab scientists to computational ones. Particularly, we expect that Antidote will become indispensable to institutions with limited resources, such as academia, not-for-profit organisations, and research institutes in developing economies.