Predicting Cancer Risk from Germline Whole-exome Sequencing Data Using a Novel Context-based Variant Aggregation...

Area of Science:

• Genetics
• Cancer Etiology
• Bioinformatics

Background:

• Genomic, nucleotide, and epigenetic contexts of somatic variants inform cancer etiology.
• Germline variant contexts show associations with oncogenic pathways, histologic subtypes, and prognosis.

Purpose of the Study:

• To determine if aggregating germline variants using meta-features improves cancer risk prediction.
• To investigate the role of rare genetic variants in cancer heritability.

Main Methods:

• Utilized germline whole-exome sequencing data from the UK Biobank.
• Developed cancer risk models for 10 cancer types using known risk variants and meta-features.
• Employed novel statistical methods to analyze rare genetic variants.

Main Results:

• Meta-features did not improve the prediction accuracy of models based on known risk variants.
• The aggregation approach did not increase statistical power for detecting signals from rare variants.

Conclusions:

• Aggregating germline variants by genomic, nucleotide, and epigenetic contexts does not currently enhance cancer risk prediction accuracy.
• Future studies incorporating whole-genome sequencing data may yield improved prediction gains.