Analysis of the genetic architecture of susceptibility to cervical cancer indicates that common SNPs explain a large proportion of the heritability

Abstract

The genetic architecture of susceptibility to cervical cancer is not well-understood. By using a genome-wide association study (GWAS) of 1034 cervical cancer patients and 3948 controls with 632668 single-nucleotide polymorphisms (SNPs), we estimated that 24.0% [standard error (SE) = 5.9%, P = 3.19×10-6] of variation in liability to cervical cancer is captured by autosomal SNPs, a bit lower than the heritability estimated from family study (27.0%), suggesting that a substantial proportion of the heritability is tagged by common SNPs. The remaining missing heritability most probably reflects incomplete linkage disequilibrium between causal variants and the genotyped SNPs. The variance explained by each chromosome is not related to its length (R2 = 0.020, P = 0.516). Published genome-wide significant variants only explain 2.1% (SE = 1.5%, P = 0) of phenotypic variance, which reveals that most of the heritability has not been detected, presumably due to small individual effects. Another 2.1% (SE = 1.1%, P = 0.013) of variation is attributable to biological pathways associated with risk of cervical cancer, supporting that pathway analysis can identify part of the hidden heritability. Except for human leukocyte antigen genes and MHC class I polypeptide-related sequence A (MICA), none of the 82 candidate genes/regions reported in other association studies contributes to the heritability of cervical cancer in our dataset. This study shows that risk of cervical cancer is influenced by many common germline genetic variants of small effects. The findings are important for further study design to identify the hiding heritability that has not yet been revealed. More susceptibility loci are yet to be found in GWASs with higher power.