This work establishes the first massive-scale atlas of drug-target affinity predictions and enables users to identify new potential targets for marketed drugs.
Background: The drug discovery process is costly, time-consuming, and lacks the detailed evaluation of drug selectivity in a comprehensive manner. Drug-repurposing, which can be used to speed-up drug development and reduce its costs, attempts to identify new targets for already approved drugs. This process involves the evaluation of a large number of drug-protein interactions and can be used as virtual counter-screening to tackle the problem of novel target sensitivity and specificity. While modern in silico methods, such as drug-target affinity predictions of large numbers of drugs made this process more feasible and scalable, to date, only few attempts have been made in creating user friendly massive-scale datasets.
Results: In this work, a comprehensive and open drug-repurposing database has been created, which explores drug-target affinify for the whole human proteome and all marketed drugs, thus covering more than 200 million drug-target pairs. The database is built, utilizing state-of-the-art deep learning-based drug-target affinity models and contains more than 4 billion predictions.
Sultanova, M., Vinogradova, E., Amantay, A., Molnár, F., Fazli, S., Sep 2023. DTA Atlas: A Massive-Scale Drug Repurposing Database. version 0.9.0. DOI: 10.5281/zenodo.8333421