PMC

Combined Analysis of Melanoma Case-Control Data from Australia, the UK, and Sweden for rs12203592 by Tumor Site, Showing the Strongest Association with Melanoma on the Trunk: OR = 1.32, p = 2.5 × 10⁻⁵

Adolescent Macular and Papular Mole Counts versus rs12203592 Genotype, Stratified on the Four Levels of Freckling Score
The association between papular counts and genotype is highly statistically significant in all cells except the first (no freckling).

Self-Reported Mole Scores in Queensland Melanoma Cases versus Age and Genotype at rs12203592 in IRF4
Bar shading represents the proportion reporting each of the four possible responses, from “none” (no shading) to “very many” (solid shading). Differences between genotypes are highly significant (see ).

Manhattan Plot of GWAS p Values for Counts of All Moles, Macular Moles, and Papular Moles
Dark points represent results for directly measured SNP genotypes, lighter points represent those for imputed SNP genotypes. The lower line represents results from linkage analysis of the same data set.

Nevus Counts versus Age and IRF4 Genotype
Lines represent cubic-polynomial-smoothed predictions from the best-fitting linear model (dotted lines are 95% confidence intervals). The three sets of lines represent Australian adolescent twins (BTNS), combined data from the UK control populations, and UK melanoma cases. Likelihood-ratio tests showed that the three UK control groups could be combined (p = 0.94).