ChemDiv has launched Kinome Profile—an advanced AI‑based binary classifier covering 50 key kinase targets.
Kinome Profile is a 3D graph neural network that leverages 3D pharmacophore features, physicochemical descriptors, and Morgan fingerprints. Pharmacophores are derived from an ensemble of 3D conformers generated with a proprietary in‑house method.
Kinome Profile enables reliable, efficient predictions of kinase activity, helping teams save time and resources in drug discovery.

Coverage (50 kinases):
AXL, BLK, BMX, BRSK1, BTK, CAMK1D, CLK2, DAPK1, DDR2, EGFR, EPHA2, EPHA3, EPHB2, ERBB4, FES, FGFR1, FGFR3, FGR, FLT3, FLT4, FYN, GRK7, GSK3A, GSK3B, HCK, IGF1R, INSR, IRAK4, KIT, LCK, MARK1, MELK, MET, NEK2, NEK7, PAK2, PHKG2, PIM1, PIM2, PIM3, PLK1, PRKX, RET, ROCK1, ROCK2, SGK2, SGK3, TBK1, TYRO3, ZAP70.

Upload up to 250 structures in a single file to generate predictions across all 50 targets.

Further details are provided in the article.