Systems Biology-Enabled Machine Learning/Artificial Intelligence to Accelerate TB Drug Discovery

Nitin Baliga of the Institute for Systems Biology in the U.S. together with Google Applied Science will combine systems biology with machine learning and artificial intelligence to accelerate the discovery of more effective and affordable treatments for tuberculosis. Tuberculosis kills 1.5 million people annually, but developing novel treatments is expensive using current methods and complicated by the different physiological states and sub-populations of the causative Mycobacterium tuberculosis. To address this, they will use a new modelling approach that leverages transcriptome data and a novel algorithm to identify more robust protein targets that are valid across bacterial states and populations. These targets will then be screened using DNA-encoded small molecule libraries (DELs), which is lower-cost than traditional high-throughput screens. Screening results will be used to train a machine learning model to identify small molecule compounds with favorable drug-like properties and high probabilities of inhibiting the target. The anti-bacterial activity of selected compounds will then be tested experimentally.