Genome-wide association studies (GWASs) aim to configure the relationship between a genotype and a phenotype through a group of cases and controls [1]. GWAS was successful in identifying many single nucleotide polymorphisms (SNPs) statistically associated with monogenic and complex diseases through the past two decades [2, 3]. However, GWASs lack the causative biological meaning in their identified variants. Moreover, the linkage disequilibrium (LD) among SNPs overshadows the causative variants [4].
The statistically significant variants (biomarkers) – that are identified from a well-conducted GWAS followed by an LD study – could be: 1) coding variants: affecting the sequence of certain gene by either: stop/loss, stop/gain, frame shift, missense mutations, or large deletions, or 2) regulatory variants: affecting gene switches or participating in posttranscriptional gene regulation. For most of the complex diseases, the bulk of the biomarkers (90%) appear to be regulatory variants – coding variants are the exception case [5].
To shed the light on the GWAS results, a fine mapping should be conducted to identify the causal genes which should lead to the biological mechanisms and the downstream pathways contributing to the disease. One of the fine mapping widespread techniques is the transcriptome-wide association study (TWAS). In case of regulatory variants, it is hard to identify the causal gene as they change the expression levels of many target genes. TWAS aims to integrate the GWAS results with the gene expression datasets to prioritize the GWAS associated variants – mostly regulatory variants - and detect the causal genes [6].
Melanoma is a typical form of cancer that affects mainly the skin as well as the iris and rectum. While melanoma represents 4% of skin cancer, it is responsible of 75% of deaths due to this type of cancer. Australia and New Zealand populations are affected by melanoma with 0.06% approximately per year. The disease is also spread in Europe and Northern America mainly among White populations. In Africans and Asians with dark skin, the spread of the disease is as low as one patient per 100,000 per year. It most often starts within the range of 40–60 years of age. Women survival rate is better than men with no known reason.
It is believed that the tendency to develop melanoma may be genetically inherited. Moreover, environmental factors, such as dense ultraviolet (UV) exposure, may excite and develop the disease in genetically susceptible children – or adults – mainly through DNA damage. Melanoma could be diagnosed by naked-eye, dermoscopy, biopsy of benign skin lesions, sequential digital dermoscopic imaging, in vivo reflectance confocal laser microscopy, computer-aided multispectral digital analysis, or electrical impedance spectroscopy [7, 8]. Brain, lungs, liver and bones are target sites for melanoma metastasis [9].
Meta-analysis GWAS have marked 54 loci that may be associated with cutaneous melanoma susceptibility. This GWAS suggested nevus formation, pigmentation and telomere maintenance as biological pathways that contribute to cutaneous melanoma [10]. Five genes (ZFP90, HEBP1, MSC, CBWD1, and RP11-383H13.1) were detected as melanoma biomarkers by TWAS using melanocyte eQTLs (expression quantitative trait loci) [11]. Fidalgo et al. identified MYO7A, WRN, SERPINB4, HRNR, and NOP10 through whole-exome sequencing (WGS) as rare variants for cutaneous melanoma in Brazilian families [12].
The top ten genes - EGFR, IL8, ICAM1, STAT1, CDK2, OAS2, MITF, PBK, CDKN3, and CXCL10 – were confirmed for association with melanoma through bioinformatic analysis in differentially expressed genes (DEGs) [13]. Logit regression and survival analysis were applied to detect the eleven feature DEGs - ZNF750, NLRP6, TGM3, KRTDAP, CAMSAP3, KRT6C, CALML5, SPRR2E, CD3G, RTP5, and FAM83C – related to the prognosis of metastatic cutaneous melanoma [14].
MDM2, MDM4, USP7, and PPM1D were confirmed in metastatic melanoma patients [15]. Spanish individuals bearing wildtype MC1R were predisposed to melanoma with a main role of HERC2 gene. ESR1 had a protective role against naevus count in female melanoma patients [16]. Walbrecq et al. suggested six proteins (AKR7A2, DDX39B, EIF3C, FARSA, PRMT5, VARS) and four microRNAs (miRNAs) (miR-210, miR-1290, miR-23a-5p, miR-23b-5p) as biomarkers for melanoma [17].
The aim of this study is to conduct a TWAS for the identification of cutaneous melanoma’s causal genes using online available GWAS summary statistics and gene expression dataset. After the identification of the causal genes, they will be related to melanoma pathology and linked to other molecular interactions and gene-disease associations by applying a functional pathway analysis and disease enrichment analysis.