Leveraging deep learning and structure-based drug repurposing for the discovery of potent Trk-A inhibitors targeting...

Nerve growth factor receptor Trk-A is essential for neurons' survival, growth, and function. The mutations in its gene can cause a disorder called CIPA, which requires the inhibition of Trk-A for preventing CIPA. Reported inhibitors of Trk-A exhibited challenges related to specificity and selectivity. This study used an integrated deep learning and structure-based drug repurposing approach to identify novel, potent, and non-toxic Trk-A inhibitors. Deep learning-based artificial neural network (ANN) models were trained and tested based on Trk-A targeting compounds' already existing bioactivity data. The trained ANN models were assessed for their performance, and a reliable model was employed to screen highly potent compounds from the FDA-approved drugs library. The screened drugs were further evaluated using molecular docking and identified gonadoliberin, caerulein, anidulafungin, and micafungin as potential Trk-A inhibitors. Molecular simulation analysis, including RMSD, RMSF, Rg, H-bonds, PCA, and MMGBSA/MMPBSA, further confirmed the stability of the selected drugs with Trk-A.