Clinical · Bioinfo & AI
Computational screening of oncogenic genetic variations in tumor suppressor proteins driving gastric cancer pathogenesis
Summary
This study employed a deep learning-based graph neural network model to identify genes and pathogenic genetic variations driving gastric cancer pathogenesis. Researchers focused on tumor suppressor proteins exhibiting both dysregulated expression and high mutation propensity. The model identified 1,886 relevant genes, with TP53, ERBB2, CDH1, and APC emerging as prominent hub genes. Missense and nonsense variants in TP53, CDH1, and APC were validated, affecting post-translational modification sites and correlating with altered patient survival. The study also highlighted EZH2 and miR-129-5p as key regulators and the dysregulation of oncogenes like CCNE1/2 and FGFR2.
Critical appraisal
Clinical. This bioinformatic analysis identifies tumor suppressor genes (TP53, CDH1, APC) and oncogenes (ERBB2, CCNE1/2, FGFR2) whose alterations are associated with altered patient survival in gastric cancer (p < 0.05). The discovery of missense and nonsense variants affecting post-translational modification sites in these genes, along with the identification of regulators like EZH2 and miR-129-5p, suggests novel prognostic and predictive biomarkers. These findings could justify early-phase clinical trials to evaluate targeted therapies based on these alterations, with a potential clinical impact in 5+ years for precise therapeutic decisions.
Bioinfo & AI. The study developed a deep learning-based graph neural network model for system-level computational screening. This model demonstrated robust performance (MSE: 0.00482 ± 0.00023 to 0.07108 ± 0.00437; R²: 0.85098 ± 0.01903 to 0.85899 ± 0.01987; AUC-ROC: 0.93095 ± 0.01758 to 0.93309 ± 0.00725) in identifying 1,886 genes exhibiting both differential expression and mutation propensity. The approach is novel in its integration of expression and mutation data, validated using cBioPortal and survival data. If deployed clinically, such a model could prioritize variants for manual review or molecular tumor board discussions, facilitating biomarker identification and patient stratification for clinical trials.
The summary and the appraisal were written by a language model from the title and abstract of the original article. A curator decided to publish it. Methodology