A novel biological function-based method for mining core genes in rare disease with limited cases

Area of Science:

• Genomics and Bioinformatics
• Rare Disease Research
• Cancer Biology

Background:

• Conventional high-throughput sequencing methods struggle with rare diseases due to small sample sizes.
• Existing methods often lack biological interpretability for gene function analysis.
• Rare subtypes like signet-ring cell carcinoma (SRCC) present unique analytical challenges.

Purpose of the Study:

• To develop a biological-function-based method for mining core genes in rare diseases with limited samples.
• To improve the biological interpretability of gene mining results.
• To provide a reliable basis for precision medicine in rare disease research.

Main Methods:

• Differential expression analysis comparing SRCC samples against broader colorectal cancer (CRC) cohorts.
• Progressive intersection of differentially expressed genes to identify initial core gene sets.
• Functional enrichment analysis and pathway integration for non-core genes.
• Secondary mining using independent samples and validation with RankProd, RRA, and an external dataset.

Main Results:

• Identified 246 initial core genes in SRCC, with secondary mining yielding 66 and 65 significant core genes from validation samples.
• Achieved high overlap between identified core genes in validation samples and across different analytical methods (RP, RRA).
• Demonstrated stable performance on both SRCC and an external mucinous adenocarcinoma dataset.

Conclusions:

• The proposed method effectively identifies biologically significant core genes in rare diseases, even with limited sample sizes.
• The approach enhances the biological interpretability of gene mining outcomes.
• This method offers a robust foundation for advancing precision medicine in rare disease research.