Six Core Gene Markers for Colorectal Cancer Malignant Transformation

Bioinformatics research from Tao He identifies six core genes (EIF2S3, GTF3A, HMGA1, HSP90AB1, PABPC1, S100A11) as markers for malignant transformation in colorectal polyps. The study constructs a ridge regression diagnostic model validated on the TCGA-COADREAD cohort and an external dataset, GSE41258. Protein and mRNA expression levels of these genes were confirmed via the UALCAN database and qRT-PCR.

Use Cases

• Validate the diagnostic model's performance using the six core genes (EIF2S3, GTF3A, HMGA1, HSP90AB1, PABPC1, S100A11) on independent cohorts like GSE41258.
• Investigate the correlation between high expression of EIF2S3 and S100A11 genes and poor prognosis in colorectal cancer patients via survival analysis.
• Compare mRNA expression levels of the six core genes between CRC cell lines (SW480, HCT116) and the normal cell line NCM460 using qRT-PCR data.
• Analyze protein expression levels of the six core genes in tumor versus normal tissues using validation data from the UALCAN database.

Strengths

• Model validation was performed on both an internal TCGA cohort and an external independent dataset (GSE41258).
• Findings are supported by multiple validation methods: bioinformatics, single-cell RNA-seq analysis, protein database queries, and laboratory qRT-PCR.
• The research specifically identifies six core genes with demonstrated diagnostic and prognostic relevance.

Limitations

• The dataset is very small (10.9 KB), containing summary results rather than raw expression data or a large patient cohort.
• Input does not specify the number of samples or rows used in the model, limiting reproducibility and scale assessment.
• Geographic and demographic details of the patient cohorts are not provided, which may limit generalizability.

Provenance

Source

Analysis derived from public datasets GSE209741, GSE161277, TCGA-COADREAD, and GSE41258.

Collection Method

Bioinformatics analysis using edgeR, WGCNA, Boruta algorithm, LASSO regression, XGBoost, and ridge regression modeling.

Freshness

Last updated March 24, 2026.