Population Substructure Has Implications in Validating Next-Generation Cancer Genomics Studies with TCGA

Area of Science:

• Genomics
• Population Genetics
• Cancer Research

Background:

• Large genetic datasets necessitate result validation using public resources like The Cancer Genome Atlas (TCGA).
• Generalizability of findings from independent or public datasets to broader populations remains a critical question.
• Understanding population genetic substructure is essential for accurate genomic study validation.

Purpose of the Study:

• To assess the generalizability of genomic findings by comparing population substructure between an independent dataset and TCGA.
• To highlight the importance of accounting for genetic differences when validating studies across different population samples.

Main Methods:

• Utilized next-generation sequencing data from endometrial and ovarian cancer patients.
• Analyzed genomic admixture using STRUCTURE and ADMIXTURE software.
• Compared genetic substructure between an independent dataset (University of Iowa) and The Cancer Genome Atlas (TCGA).

Main Results:

• Identified one subpopulation in the independent dataset.
• Detected 4-6 subpopulations within the TCGA dataset.
• Demonstrated significant differences in genetic substructure between the University of Iowa and TCGA populations.

Conclusions:

• The genetic substructure of TCGA and independent population samples can differ substantially.
• Validation of genomic studies requires awareness and correction for background genetic substructure.
• Ensuring population representativeness is crucial for the reliability of cancer genomics research.