Joint effect of multiple common SNPs predicts melanoma susceptibility.

Published online

Journal Article

Single genetic variants discovered so far have been only weakly associated with melanoma. This study aims to use multiple single nucleotide polymorphisms (SNPs) jointly to obtain a larger genetic effect and to improve the predictive value of a conventional phenotypic model. We analyzed 11 SNPs that were associated with melanoma risk in previous studies and were genotyped in MD Anderson Cancer Center (MDACC) and Harvard Medical School investigations. Participants with ≥15 risk alleles were 5-fold more likely to have melanoma compared to those carrying ≤6. Compared to a model using the most significant single variant rs12913832, the increase in predictive value for the model using a polygenic risk score (PRS) comprised of 11 SNPs was 0.07(95% CI, 0.05-0.07). The overall predictive value of the PRS together with conventional phenotypic factors in the MDACC population was 0.69 (95% CI, 0.64-0.69). PRS significantly improved the risk prediction and reclassification in melanoma as compared with the conventional model. Our study suggests that a polygenic profile can improve the predictive value of an individual gene polymorphism and may be able to significantly improve the predictive value beyond conventional phenotypic melanoma risk factors.

Full Text

Duke Authors

Cited Authors

  • Fang, S; Han, J; Zhang, M; Wang, L-E; Wei, Q; Amos, CI; Lee, JE

Published Date

  • 2013

Published In

Volume / Issue

  • 8 / 12

Start / End Page

  • e85642 -

PubMed ID

  • 24392023

Pubmed Central ID

  • 24392023

Electronic International Standard Serial Number (EISSN)

  • 1932-6203

Digital Object Identifier (DOI)

  • 10.1371/journal.pone.0085642

Language

  • eng

Conference Location

  • United States