Machine learning-based screening and molecular simulations for discovering novel PARP-1 inhibitors targeting DNA repair mechanisms for breast cancer therapy.

Abstract

Cancer remains one of the leading causes of death worldwide, with the rising incidence of breast cancer being a significant public health concern. Poly (ADP-ribose) polymerase-1 (PARP-1) has emerged as a promising therapeutic target for breast cancer treatment due to its crucial role in DNA repair. This study aimed to discover novel, targeted, and non-toxic PARP-1 inhibitors using an integrated approach that combines machine learning-based screening, molecular docking simulations, and quantum mechanical calculations. We trained a widely used machine learning models, Random Forest, using bioactivity data from known PARP-1 inhibitors. After evaluating the performance, it was used to screen an FDA-approved drug library, successfully identifying Atazanavir, Brexpiprazole, Raltegravir, and Nisoldipine as potential PARP-1 inhibitors. These compounds were further validated through molecular docking and all-atom molecular dynamics simulations, highlighting their potential for breast cancer therapy. The binding free energies indicated that Atazanavir at - 41.86 kJ/mol and Brexpiprazole at - 45.44 kJ/mol exhibited superior binding affinity compared to the control drug at - 30.42 kJ/mol, highlighting their promise as candidates for breast cancer therapy. Subsequent optimized geometries and electron density mappings of the two molecular structures revealed a Gibbs free energy of - 2334.610 Ha for the first molecule and - 1682.278316 Ha for the second, confirming enhanced stability compared to the standard drug. This study not only highlights the efficacy of machine learning in drug discovery but also underscores the importance of quantum mechanics in validating molecular stability, setting a robust foundation for future pharmacological explorations. Additionally, this approach could revolutionize the drug repurposing process by significantly reducing the time and cost associated with traditional drug development methods. Our results establish a promising basis for subsequent research aimed at optimizing these PARP-1 inhibitors for clinical use, potentially offering more effective treatment options for breast cancer patients.