Pangenome-based somatic mutation landscape of early- and late-onset colorectal cancer in 99 Korean patients

Colorectal cancer (CRC) is the third most prevalent cancer type worldwide. Although overall CRC incidence has declined due to improved screening, early-onset CRC (EOCRC) in patients under 50 years old is rapidly increasing globally, including in Korea, unlike the decreasing trend of late-onset CRC (LOCRC). Several studies have attempted to identify EOCRC-specific mutations; however, most were limited to whole-exome sequencing and thus failed to capture CRC-related non-coding variants comprehensively. Moreover, previous mutation-profiling studies relied on the GRCh38 reference genome, a linear and largely European-biased assembly that lacks representation of human genetic diversity, fundamentally limiting the discovery of novel mutations. To address these limitations, we performed whole-genome sequencing (WGS) of paired tumor and matched blood samples from 49 EOCRC and 50 LOCRC Korean patients. Somatic mutations were identified using the graph-based Human Pangenome Reference (Human Pangenome Reference Consortium, released in 2023), which integrates genomic diversity from 49 individuals worldwide. Each sample produced 1.409 billion reads with an average coverage of 37.7×. EOCRC and LOCRC carried an average of 1,673 and 801 somatic mutations in coding regions and 236,610 and 127,616 in non-coding regions-21% and 43% higher than GRCh38-based results. Ten EOCRC-specific coding genes were identified, including LMTK3, GALNT11, and NPHP1, with LMTK3 also reported in a previous Chinese study. In non-coding regions, 32 EOCRC-specific genomic regions were detected, including PRKAG2-GALNTL5, MRGPRF-TPCN2, and ACTRT2-TTC34. Among these, GALNT11, MRGPRF-TPCN2, and ACTRT2-TTC34 contained mutations significantly associated with EOCRC patient survival. Although further expression-level validation is warranted, we propose these three regions as promising genome-level candidates for EOCRC diagnosis and therapy. Additionally, all WGS data have been deposited in the European Genome-phenome Archive (EGA; EGAS50000000544) and are expected to provide a valuable resource for researchers investigating CRC or EOCRC at the genomic level.