Integrated network toxicology, machine learning algorithms and TMT proteomics reveal the mechanism of 18β glycyrrhetinic acid against gastric cancer

TL;DR

This study uses toxicology and machine learning to identify how 18β-GRA may treat gastric cancer by finding key biomarkers and their interactions.

Contribution

A novel integration of network toxicology, machine learning, and proteomics to identify biomarkers for 18β-GRA's anti-gastric cancer effects.

Findings

12 overlapping targets were identified between WGCNA and TMT proteomics analyses.

Three candidate biomarkers (IGF2BP3, KRT6B, NEDD4L) were selected using machine learning algorithms.

NEDD4L is proposed as a key biomarker for 18β-GRA, with SCN5A and EGR1 as potential regulatory proteins.

Abstract

The purpose of this paper is to explore the mechanism of 18β glycyrrhetinic acid (18β-GRA) in treating gastric cancer. Firstly, the toxicological effects of 18β-GRA were predicted using the ProTox3.0 database. Then, candidate biomarkers for the anti-gastric cancer of 18β-GRA were screened using the weighted gene co-expression network analysis (WGCNA), the least absolute shrinkage and selection operator (LASSO), the support vector machine (SVM), the random forest algorithm combined with the TMT proteomics methods. Additionally, we explored the potential upstream transcription factors and downstream interacting proteins of the biomarkers. The WGCNA method yielded 269 targets, while TMT proteomics analysis identified 6,273 genes. Among these, 12 targets were identical. Using LASSO, SVM, and random forest algorithms, three candidate markers were identified: insulin-like growth factor 2 mRNA binding protein 3 (IGF2BP3), keratin 6B (KRT6B), and E3 ubiquitin-protein ligase NEDD4-like (NEDD4L). Based on molecular docking and molecular dynamics results, NEDD4L is believed to be a 18β-GRA biomarker, while sodium channel protein type 5 subunit alpha (SCN5A) and early growth response protein 1 (EGR1) are the potential upstream and downstream regulatory proteins, respectively. These findings provide a theoretical basis for future experimental verification.

Diagram illustrating a multi-step analytical process, including sections on network toxicology, disease targets, drug targets, screening biomarkers, and molecular mechanisms. Each section contains diagrams, graphs, and tables showing the workflow and findings for drug toxicity, WGCNA, proteomics, biomarker screening, and mechanism analysis.

Figures