Identification\nof Potent CDK9 Inhibitors with Novel\nSkeletons via Virtual Screening, Biological Evaluation, and Molecular\nDynamics Simulation

Cyclin-dependent kinase 9 (CDK9) is a pivotal therapeutic\ntarget\nfor acute myeloid leukemia (AML), a hematologic malignancy characterized\nby limited effective treatments. In this study, we introduced an innovative\nvirtual screening strategy combines machine learning models with molecular\ndocking techniques. We constructed a comprehensive CDK9 data set,\nfollowed by applying molecular fingerprints and selecting machine\nlearning models, particularly random forest models, for virtual screening.\nMolecular docking, combined with similarity and drug-likeness screening\nwas used to ensure scaffold diversity from existing CDK9 inhibitors.\nThis integrative strategy led to the identification of a new class\nof CDK9 inhibitors with scaffolds distinct from traditional CDK9 molecules.\nHere, Compound 2 demonstrated significant inhibitory activity against\nCDK9 and reduced cell viability effectively in hematologic tumor cell\nlines. Additional validation using molecular dynamics simulation clarified\nthe binding mechanism of the inhibitor-protein complex. These findings\nhighlight the effectiveness of our hybrid virtual screening approach\nin drug discovery, offering a promising new scaffold for CDK9 inhibitors.