Off-target mapping enhances selectivity of machine learning-predicted CK2 inhibitors
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
A key challenge in drug development is identification of druggable targets, the modulation of which attenuates disease progression, while avoiding inhibition of proteins that lead to dose-limiting toxicities. Here, we investigate a drug target Casein kinase 2 (CK2) - a serine/threonine kinase implicated in cancer, for which existing molecules have so far failed clinical trials. Using molecular and pharmacoepidemiology approaches, we show that molecules targeting CDK kinase family members CDK1/2/7/9 - such as the existing CK2 inhibitors - have a higher risk to induce adverse effects or fail in clinical trials. Based on this finding, we establish a machine learning assisted pipeline to redesign more specific and allosteric lead compounds against CK2, with more selective on-target binding and favourable off-target profile. Importantly, we show that such design is possible via machine learning powered, docking assisted discovery pipeline, when standard ML algorithms were combined with an error prediction model. In conclusion, our study reports a simple yet efficient machine learning-powered drug discovery pipeline and novel submicromolar CK2 inhibitors targeted. Importantly, our prediction pipeline was able to achieve a 90% hitrate, significantly reducing the need for subsequent wet-lab validation.