Computational Modeling and Virtual Screening of Small Molecule Inhibitors Targeting BAIAP2 in Medulloblastoma

Tejaswini Tushar Oak

Mentor: Dr. Nagi Ayad, The Ayad Lab, Department of Oncology, Lombardi Comprehensive Cancer Center, Georgetown University Medical Center.

Date/Time: August 22nd, 2025 at 9:15 AM.

Abstract: Medulloblastoma is a highly malignant pediatric brain tumor with limited long-term treatment outcomes due to recurrence and therapy-induced neurotoxicity. BAIAP2 is a key gene overexpressed in medulloblastoma, where it promotes tumor cell migration and invasion through interaction with CDC42, a Rho GTPase involved in cytoskeletal remodeling. BAIAP2 and CDC42 have not been previously characterized and The Ayad Lab was the first one to do it.

In this project, I used structure-based virtual screening to identify novel small molecule inhibitors of BAIAP2. The 3D structures of three BAIAP2 domains were used as the receptor for molecular docking simulations with a curated library of candidate ligands. Using AutoDock Vina and Bash scripting, I automated the docking pipeline, including receptor and ligand preparation, high-throughput docking of ~27,000 ligands, and extraction of binding affinities. The top-scoring ligands were visualized using UCSF Chimera to examine predicted binding modes and their potential to disrupt protein-protein interactions. Two small molecules were identified that bind strongly to BAIAP2 and one of them binds strongly to the CRIB-PR domain through which BAIAP2 binds to CDC42. This work provides a computational framework for identifying lead compounds targeting BAIAP2 and lays the groundwork for future cheminformatics-based prioritization or experimental validation. The approach integrates bioinformatics, structural biology, and drug discovery to address a critical therapeutic gap in medulloblastoma.