Recently, the research team led by Dr. Hu Chengbin from the School's Interdisciplinary Center, in collaboration with the Guangdong Province Hospital for Occupational Disease Prevention and Treatment, published a research paper in the international toxicology journal Toxicology (Impact Factor: 4.6, Q2), a high-ranking journal in the field. The study focuses on the hepatotoxicity mechanism of the environmental pollutant 6PPD-quinone. The molecular targets and mechanisms underlying its hepatotoxic effects had long been unclear. The team innovatively employed a purely computational research paradigm integrating "network toxicology + molecular docking/molecular dynamics simulation (MDS) + machine learning," efficiently filling this knowledge gap. The manuscript was accepted within just two months from submission, offering a practical new pathway for the toxicity analysis of environmental pollutants.

The study first screened and cross-verified 62 key hepatotoxicity targets of 6PPD-quinone using the Comparative Toxicogenomics Database (CTD) and SwissTargetPrediction database. By constructing a protein-protein interaction network with STRING and conducting pathway enrichment analysis via Metascape, the team found these targets were primarily enriched in the PI3K-Akt/MAPK signaling pathway (core targets: EGFR/AKT1) and xenobiotic metabolism pathways (core targets: CYP2B6/CYP3A4). This revealed a dual mechanism for hepatotoxicity induction through "signaling disruption + metabolic dysregulation." Subsequently, molecular docking verification using AutoDock Vina showed that 6PPD-quinone achieved binding energies of -7.65 kcal/mol with EGFR and -7.62 kcal/mol with CYP2B6, with all core target binding energies ≤ -5.0 kcal/mol. Further confirmation through a 100ns MDS performed with GROMACS demonstrated stable binding between 6PPD-quinone and EGFR (key residues CYS797/MET793 maintained hydrogen bond occupancies of 38%/24%, respectively). Finally, utilizing datasets built from the PubChem and ChEMBL databases, the team developed classification models for target modulators using Random Forest and Support Vector Machine (SVM). The SVM model predicted an 88% probability of 6PPD-quinone inhibiting EGFR (AUC 0.93-0.99), and the Random Forest model achieved an accuracy of 0.91 in distinguishing target inhibitors/agonists. Notably, CYP3A4 was predicted as a potential agonistic target, offering a new direction for research into detoxification mechanisms. The paper is titled "EGFR and CYP signaling disruption underlies 6PPD-quinone hepatotoxicity: Insights from a network and machine learning approach" and fully embodies the innovative value of interdisciplinary "Medicine + Engineering" research.

This achievement not only represents the first systematic elucidation of the hepatotoxicity mechanism of 6PPD-quinone but, more importantly, the research paradigm of "network toxicology + structural biology + machine learning" demonstrates significant advantages in efficiency and low cost. It can be conducted without complex experimental setups, making it particularly suitable for teams lacking extensive laboratory platforms. Simultaneously, as a representative achievement in the field of interdisciplinary research, it showcases the team's technical prowess in addressing environmental and health challenges and provides a practical case study for similar interdisciplinary research endeavors.

Link to the publication: https://doi.org/10.1016/j.tox.2025.154235