The advent of “big data” from medical field research helps bring chemical medical countermeasure (cMCM) solutions to the warfighter faster and at a reduced cost as cMCM development is more efficiently focused on treatments highly likely to be effective.
The Defense Threat Reduction Agency’s (DTRA) Chemical and Biological Technologies Department in its role as the Joint Science and Technology Office (JSTO) uses chemical informatics, or “chemoinformatics,” to mine a Department of Defense (DoD) repository of this composite data. Chemoinformatics focuses on storing, indexing, searching, retrieving, and applying information about chemical compounds. Researchers receiving DTRA-JSTO investments perform together to streamline the formatting, standardization, capture, transfer mechanisms, analysis, storage, and querying processes associated with big data generated by various performers’ cMCM discovery campaigns.
Advances in biotechnology that enable more drug discovery data than ever before to be generated at a fixed cost also compound challenges associated with big data,* but the following technologies in drug development aid performers to more rapidly create effective candidate cMCMs:
- High-throughput screening (HTS)—an automated method to rapidly test hundreds to thousands of predictions about the behavior of different compounds
- Artificial intelligence (AI)—intelligence as demonstrated by machines, unlike the natural intelligence displayed by humans and animals
- Machine learning (ML)—an application of AI that provides systems the ability to automatically learn and improve from experience without being explicitly programmed
While ultra-HTS of a chemical library against one target can produce millions of data points, multiple HTS efforts by a single performer can result in big data sets so large that they cannot be processed by conventional methods such as using reports or spreadsheets.
This coordination and collaboration between performers will overcome known issues associated with applying AI/ML to big data, and fully utilizing big data will:
- Enable earlier safety and efficacy assessments of candidate cMCMs
- Enhance candidate cMCM hit-to-lead down-select decisions—the pairing down of identified candidate cMCMs to a smaller number of leading candidates based on safety and efficacy data
- Reduce redundancy of the unintentional duplication of testing or screening of the same candidate cMCMs by different performers
- Increase the efficiency of cMCM development
This system will deliver novel candidate cMCMs to the warfighter faster and potentially save lives. For example, DTRA-JSTO research performed at Alchem Laboratories Corporation is using ultra-HTS to screen 370,000 compounds in the National Institutes of Health Molecular Libraries Small Molecule Repository library for several receptors that can be targeted by cMCMs to protect against nerve agent poisoning.
Anticipating the increased informatics and computational requirements needed to curate, prioritize, store, analyze, and manage this data, DTRA-JSTO also funds researchers at the Biotechnology High Performance Computing Software Applications Institute (BHSAI) of the U.S. Army Medical Research and Development Command Telemedicine and Advanced Technology Research Center to provide AI/ML and computational chemistry expertise in support of Alchem and other similar efforts in the cMCM portfolios.
BHSAI’s current effort involves creating an HTS database and deploying AI-based algorithms to mine that database and support hit prioritization integrated into BHSAI’s access-controlled Chemoinformatics System. This central repository will capture, store, and visualize datatypes from chemicals in the screening libraries and results from the HTS biological assays generated from DTRA-JSTO funded performers across the cMCM Enabling Science portfolio.
The HTS database interfaces with existing and new in-house BHSAI computational tools to predict features of candidate cMCMs that are important for drug development, such as standard absorption, distribution, metabolism, excretion, and toxicity parameters.
The HTS data will also be used as training sets to develop deep-learning algorithms to identify which chemicals in a screening library fit the profile of a good drug: maximum efficacy with minimal safety concerns. The data produced by HTS efforts will help to train new ML algorithms.
The BHSAI DoD repository for the central collection, standardization, storage, and AI/ML-based analysis of HTS candidate cMCM data will enable DTRA-JSTO to make the most out of combining AI/ML and big data to more rapidly create effective cMCM for the warfighter at a reduced cost.
*.Brown, N., Cambruzzi, J., Cox, P. J., Davies, M., Dunbar, J., Plumbley, D., Sellwood, M. A., Sim, A., Williams-Jones, B. I., Zwierzyna, M., & Sheppard, D. W. (2018). Big Data in Drug Discovery. Progress in medicinal chemistry, 57(1), 277–356. https://doi.org/10.1016/bs.pmch.2017.12.003
POC: Katherine Mann, Ph.D., katherine.m.mann2.civ@mail.mil
| Date Taken: |
06.23.2021 |
| Date Posted: |
06.23.2021 14:31 |
| Story ID: |
399551 |
| Location: |
FT. BELVOIR, VA, US |
| Web Views: |
340 |
| Downloads: |
0 |
PUBLIC DOMAIN
This work, Talking the Same Talk, must comply with the restrictions shown on https://www.dvidshub.net/about/copyright.
Subscribe
LEAVE A COMMENT