Differential expression analysis for individual cancer samples based on robust within-sample relative gene expression...

Area of Science:

• Genomics
• Bioinformatics
• Cancer Research

Background:

• Gene expression relative orderings (REOs) are stable within normal tissues but reversed in cancer.
• Existing methods for detecting differentially expressed genes (DEGs) can be platform-specific.

Purpose of the Study:

• To develop and validate a cross-platform algorithm (RankComp) for detecting DEGs using stable REOs.
• To assess the accuracy and utility of RankComp in identifying cancer-specific gene expression patterns.

Main Methods:

• Utilized 461 normal lung tissue samples across four platforms to identify stable REOs.
• Applied RankComp to detect DEGs in paired lung and colorectal cancer samples.
• Performed individualized pathway analysis on detected DEGs.

Main Results:

• Identified millions of stable REOs in normal lung tissue consistently detected across platforms.
• RankComp achieved high precision (94-96%) in detecting DEGs in lung cancer datasets.
• Uncovered common and subtype-specific functional mechanisms in lung and colorectal cancers.

Conclusions:

• Stable, cross-platform REOs enable accurate, individualized DEG detection in cancer.
• RankComp offers a robust approach for dissecting cancer heterogeneity.
• This method is applicable across different platforms and cancer types.