Free energy perturbation-assisted identification of checkpoint kinase 1 targeted small-molecule inhibitors for cancer therapy using AI- and physics-driven molecular design.

Checkpoint kinase 1 (Chk1) is a vital protein that preserves genomic integrity by regulating cellular responses to DNA damage and replication stress. Because of its key role in cell cycle control, Chk1 has become an attractive target for cancer therapy. Advanced computational methods were used to identify new Chk1 inhibitors, combining REINVENT4 for de novo molecule creation and DeLA-Drug for analogue design. A library of potential molecules was developed from known active compounds using the Mol2Mol generator in REINVENT4. Drug-like properties were verified through ADMET analysis with pkCSM. Three promising candidates, namely, CHD1, CHD2, and CHD3, were selected based on their superior binding free energies (ΔGbind values of -8.92, -6.32, and -5.25 kcal/mol, respectively) compared to the co-crystal ligand 5CV (-4.92 kcal/mol). These candidates were further assessed using molecular dynamics (MD) simulations and MM-GBSA binding free-energy calculations, and the best compound was also subjected to absolute free-energy perturbation (FEP) analysis. The selected compounds were predicted to exhibit high conformational diversity and strong dynamic stability, as demonstrated by free-energy landscape and principal component analyses. The binding disrupted the intramolecular hydrogen-bond network in Chk1 and encouraged the formation of new hydrogen bonds. These results provide a strong foundation for optimizing and developing new Chk1 inhibitors as potential anticancer agents.