Sample size required for prognostic genes analysis in colorectal cancer

Area of Science:

• Oncology
• Genomics
• Bioinformatics

Background:

• Colorectal cancer (CRC) research often lacks overlapping prognostic genes across studies.
• Investigating the influence of sample size on prognostic gene discovery in CRC is crucial.

Purpose of the Study:

• To analyze the effect of sample size on prognostic gene identification in colorectal cancer.
• To compare the impact of whole-exome sequencing (WES) versus targeted sequencing on prognostic gene discovery.

Main Methods:

• Utilized TCGA PanCancer (418 CRC cases, WES) and MSK (931 CRC cases, targeted sequencing) cohorts.
• Employed random resampling 200 times per sample size level to assess prognostic gene analysis.
• Analyzed prognostic gene identification across different colorectal cancer stages and mutation statuses.

Main Results:

• Prognostic gene identification increased with sample size in a power-law (WES) and linear (targeted sequencing) fashion for stages III/IV CRC.
• Approximately 550 cases were needed for stage IV CRC targeted sequencing to plateau gene discovery.
• WES data yielded more prognostic genes but with lower reliability compared to targeted sequencing at equivalent sample sizes.

Conclusions:

• Demonstrates a clear relationship between sample size and the number/reliability of prognostic genes in colorectal cancer.
• Highlights how sequencing methodology (WES vs. targeted) influences prognostic gene discovery.
• Provides insights for optimizing trial design in prognostic genetic analysis for CRC.