Triple-Action Therapy: Combining Machine Learning, Docking, and Dynamics to Combat BRCA1-Mutated Breast Cancer

Breast cancer dominates women's mortality, and among other factors, mutations in the BRCA1 gene are significant risk factors. Several approaches are followed to treat the BRCA1 affected cancer patients. However, specific BRCA1 inhibitors are not available till date due to its structural complexity. In addition, there are several limitations associated with the existing drugs used to treat BRCA1-related breast cancer and some side effects. The side effects include symptoms such as hot flashes, joint pain, nausea, fatigue, hair loss, diarrhea, chills, fever, and others. Therefore, advanced approaches needed that can overcome all the limitations and side effects of the current inhibitors. In this study, we adopted a multistep approach to identify potential inhibitors for BRCA1-mutated breast cancer. We used our developed machine learning models to screen potential inhibitors. Molecular docking approach was carried out for the screened hit compounds with BRCA1 and its mutated forms. Two ligands, beta-amyrin and Narirutin, has shown significant performance in multiple scoring schemes such as molecular docking and RF score calculations. Molecular dynamics simulations demonstrated the stability of the complexes formed by beta-amyrin and Narirutin with BRCA1, with lower RMSD values and less RMSF fluctuations at the binding site locations. Principal component analysis (PCA) and free energy landscape (FEL) further confirmed the compactness and favorable binding of beta-Amyrin and Narirutin to BRCA1. These findings suggest that beta-amyrin and Narirutin have potential as therapeutic agents against BRCA1-mutated breast cancer.