Large-scale identification of PARP7 inhibitors via computational modeling and simulation

Abstract

Poly(ADP-ribose) polymerases (PARPs) are attractive therapeutic targets for cancer. This study focuses on PARP7, a mono(ADP-ribose) polymerase with emerging potential in cancer therapy due to its roles in immune response and tumorigenesis. Using computational modeling, we screened millions of compounds through molecular docking, machine learning based on molecular fingerprints, molecular dynamics (MD) simulations, and ADME profiling. We identified promising PARP7 inhibitor candidates exhibiting higher binding affinity than NAD⁺ and comparable affinity to RBN2397, with favorable binding energies and pharmacological properties. MD simulations confirmed complex stability, while interaction analysis revealed key binding residues including conserved residues (Y564/H532) and hydrophobic residues (F575/I542). In silico ADME predictions indicated favorable drug-like properties and pharmacokinetic profiles. This work establishes a foundation for developing novel PARP7 inhibitors, offering new therapeutic strategies for cancer.