Overlapping genetic architecture between Parkinson disease and melanoma

Abstract

Epidemiologic studies have reported inconsistent results regarding an association between Parkinson disease (PD) and cutaneous melanoma (melanoma). Identifying shared genetic architecture between these diseases can support epidemiologic findings and identify common risk genes and biological pathways. Here, we apply polygenic, linkage disequilibrium-informed methods to the largest available case-control, genome-wide association study summary statistic data for melanoma and PD. We identify positive and significant genetic correlation (correlation: 0.17, 95% CI 0.10-0.24; P = 4.09 × 10^-06) between melanoma and PD. We further demonstrate melanoma and PD-inferred gene expression to overlap across tissues (correlation: 0.14, 95% CI 0.06 to 0.22; P = 7.87 × 10^-04) and highlight seven genes including PIEZO1, TRAPPC2L, and SOX6 as potential mediators of the genetic correlation between melanoma and PD. These findings demonstrate specific, shared genetic architecture between PD and melanoma that manifests at the level of gene expression.

Keywords: Genetic correlation; Melanoma; Parkinson disease; Polygenic; Shared genetic architecture; TWAS.

Figure 1.. GNOVA Genetic Correlation Results for Parkinson Disease and Melanoma GWAS Summary Statistic Datasets

Forest plot of genetic correlation between melanoma and the individual and meta-analyzed Parkinson disease datasets (Tables 3-4). Box size indicates the effective sample size (N_eff = 4/(1/N_cases+1/N_controls)). The three independent PD datasets are Nalls2014 (Nalls et al., 2014[64]); Chang2017 (Chang et al., 2017[13]); Nalls2019 (Nalls et al., 2019[63]). METAPD is an inverse-variance-weighted meta-analysis of the three independent Parkinson disease summary statistic datasets.

Figure 2.. Parkinson Disease (PD) and Melanoma Tissue-specific, Disease-inferred Gene Expression Profile Correlation

PD and Melanoma disease-inferred gene expression profile correlation at the level of 48 specific tissues included in the GTEx v7 reference panel (Table 5). Disease-inferred gene expression profiles were generated from the processed melanoma and METAPD summary statistics using FUSION/TWAS software and correlation between these profiles was estimated using RHOGE software. METAPD is an inverse-variance-weighted meta-analysis of the three independent Parkinson disease summary statistic datasets. The red dashed line demarks the multiple test corrected P threshold of 1.04 × 10⁻⁰³ (0.05 / 48) while the blue dotted line demarks the nominal threshold, P = 0.05.

Figure 3.. Cross-tissue eGenes Associated with Both Parkinson Disease (PD) and Melanoma.

Conjunction plot of the cross-tissue PD and melanoma eGene −log₁₀ P values. We generated cross-tissue eGene-disease results (Supplementary Tables 3-4, online resource) from the processed melanoma and METAPD summary statistics using UTMOST software. METAPD is an inverse-variance-weighted meta-analysis of the three independent Parkinson disease summary statistic datasets. The red dashed lines demark the false discovery rate (FDR) threshold of 0.05. Labels and lines indicate eGenes associated with both PD and melanoma under the FDR threshold.