Identification of Colorectal Cancer-Related RNA Markers from Whole Blood Using Integrated Bioinformatics Analysis.

Area of Science:

• Molecular diagnostics
• Genomics and transcriptomics
• Cancer biomarker discovery

Background:

• Current colorectal cancer (CRC) screening relies on DNA biomarkers, often detecting later-stage tumors.
• Whole-blood transcriptomic profiling offers a complementary approach by capturing systemic immune responses and tumor-host interactions.
• Clinically validated whole-blood transcriptomic signatures for early CRC detection are limited.

Purpose of the Study:

• To develop and validate a whole-blood RNA-based biomarker signature for early colorectal cancer (CRC) detection.
• To establish a robust analytical framework for reproducible whole-blood RNA biomarker discovery.

Main Methods:

• Integrated multi-cohort transcriptomic data (GEO datasets GSE164191, GSE11545) to identify differentially expressed genes (DEGs).
• Applied multi-layer biological filtering (PPI networks, GWAS, eQTLs, DigSeE, CoReCG) to refine DEGs into high-confidence transcripts (WB-PADs).
• Validated candidate transcripts using RNA-seq on a Korean cohort and RT-qPCR on an independent clinical cohort (106 CRC, 123 controls).

Main Results:

• A five-transcript signature (DLG5, CD177, SH2D1B, NQO2, KRT73) was identified and validated.
• The signature achieved an AUC of 0.952 in distinguishing CRC patients from healthy controls.
• High accuracy was maintained for early-stage CRC (Stage I-II: AUC 0.929), with high sensitivity at specific screening thresholds.

Conclusions:

• A robust framework for whole-blood RNA biomarker discovery was established.
• The five-gene signature demonstrates significant potential for minimally invasive and early colorectal cancer detection.
• This transcriptomic approach offers a promising complementary strategy to existing CRC screening methods.