We use machine learning to address a specific problem in drug discovery, i.e. predicting protein-small molecule binding affinities, which has implications for hit finding, potency, selectivity, and toxicity. For us, machine learning is a way to reduce the number of compounds that need to be physically screened and to screen more compounds than ever before to find new and unique scaffolds.
You can read more about how it works here: https://www.atomwise.com/our-technology/.
Here is some examples of success in this area: https://www.biocentury.com/bc-extra/preclinical-news/2019-04-17/atomwise-unveils-chagas-disease-compounds-aims-program