Unveiling key drivers of hepatocellular carcinoma: a synergistic approach with network pharmacology, machine learning-driven ligand discovery and dynamic simulations.

Abstract

Hepatocellular carcinoma (HCC) ranks fourth in cancer-related mortality worldwide. This study aims to uncover the genes and pathways involved in HCC through network pharmacology (NP) and to discover potential drugs via machine learning (ML)-based ligand screening. Additionally, toxicity prediction, molecular docking, and molecular dynamics (MD) simulations were conducted. NP study identified key genes related to HCC, particularly the enzymes AKT1 and GSK3β. Pathway analysis revealed that crucial pathways like PI3K-AKT and WNT signalling play pivotal roles in HCC progression. Using ML, potential inhibitors for AKT1 and GSK3β were identified, including CHEMBL2177361 and CHEMBL403354 for AKT1, and CHEMBL3652546 and CHEMBL4641631 for GSK3β. post-MD analyses, including RMSD, 2D-RMSD, RMSD cluster, RMSF, PCA, DCCM, residence time analysis, diffusion coefficient, autoencoder-based dimensionality reduction, FEL and MM/GBSA were performed to understand the protein-ligand interactions. The present study reveals the stable interactions of the inhibitors with AKT1 and GSK3β. The binding free energies of all the four complexes were -39.9, -46.8, -41.6, and -45.9 kcal/mol, respectively. This research provides novel insights into the genes and pathways involved in the progression and pathogenesis of HCC using bioinformatics tools. Furthermore, ML-based virtual screening identified potent inhibitors against the target proteins of HCC, such as AKT1 and GSK3β.