Integrative Mendelian Randomization and Pathomics Analysis Using Expression Quantitative Trait Loci and Genome-Wide...

Area of Science:

• Oncology
• Genetics
• Bioinformatics

Background:

• Mismatch repair (MMR) genes are linked to stomach adenocarcinoma (STAD).
• Understanding the causal role and prognostic value of MMR genes in STAD is crucial.
• Predicting gene expression from histopathology could offer new diagnostic avenues.

Purpose of the Study:

• To investigate the causal relationship between MMR genes and gastric cancer (GC).
• To evaluate the prognostic significance of MMR genes, particularly MutS homolog 2 (MSH2), in STAD.
• To develop a histopathology-based model for predicting MSH2 expression.

Main Methods:

• Mendelian randomization (MR) analysis using IEU OpenGWAS data to assess causal links between MMR genes and GC.
• Survival analysis to determine the prognostic relevance of MMR genes.
• A random forest model trained on TCGA histopathology images to predict MSH2 expression.
• Exploration of biological insights through pathomics, GSEA, immune infiltration, and tumor mutational burden (TMB).

Main Results:

• MR analysis identified MLH1 and PMS2 as risk genes, while MSH2 demonstrated a protective effect in GC.
• MSH2 was confirmed as an independent protective factor for STAD (HR=0.690, p<0.05).
• The pathomics model achieved an AUC of 0.811 for predicting MSH2 expression.
• High-survival-probability groups exhibited higher TMB and TP53 mutation frequencies.

Conclusions:

• MMR genes, particularly MSH2, are critical in STAD pathogenesis and patient prognosis.
• An image-based model effectively predicts MSH2 expression, highlighting the potential of integrating genomic and histopathologic data.
• This approach supports personalized gastric cancer care through combined data analysis.