GWAS with AHP based SNP prioritization approach to identify SNP biomarkers for Alzheimer's disease

Genome wide association studies (GWAS) is defined as search for biological variance associated with certain phenotypes and diseases among individuals in a population depending on statistical analysis. Combined p-value approach has been introduced recently and is defined as the second-wave GWAS. It helps mapping of significant SNPs to genes and pathways to evaluate SNP-gene-disease associations. Identification of enriched genes and pathways significantly associated with diseases can be performed via this approach. The major bottleneck of current standard GWAS approaches is the prioritization of statistically significant results. Our group has recently developed a novel Analytical Hierarchical Process (AHP) based on a structured SNP prioritization algorithm. SNPs are scored according to their biological relevance in terms of their genomic location and functional consequence, evolutionary conservation, and gene-disease association. The recently developed METU-SNP application integrates GWAS, combined p-value while utilizing AHP based SNP prioritization algorithms. Combined p-value and AHP prioritization approach for GWAS of Alzheimer's Disease (AD) has been utilized for the SNP-disease association of AD for the first the time in this study with METU-SNP software. The results from the analysis of two different sets of AD genotyping data with the newly proposed AHP based prioritization yield promising results for both datasets. For the ADNI data, all the top 100 SNPs according to AHP scoring map to OMIM associated genes and 18 of them map to AD linked genes. For the GenADA data, all the top 100 SNPs according to AHP scoring map to OMIM associated genes and 37 of them map to AD linked genes. Glycolysis and gluconeogenesis, leukocyte migration, axon guidance, actin filament polymerization, cell adhesion, DNA fragmentation during apoptosis, fatty acid metabolism, and negative regulation of cell proliferation are common pathways residing at top 100 pathways according to combined p-value for pathways that are observed in GWAS results of both data sets. GWAS of both data with METU-SNP confirms the literature for AD associated genes; A2M, ABCA1, ACE, APOA1, APP, CHRNA7, IL1A, LDLR, LPL, MPO, PTGS2, SORL1. rs3781835 at SORL1, rs4343, and rs4351 at ACE1 are SNPs with high AHP scores are also listed to be AD associated at PharmGKB database. Moreover, CT and TT genotype of rs6313 at HTR2A gene indicates resistance to the treatment with antipsychotic drugs for AD patients presenting delusional symptoms. As presented here METU-SNP is a powerful tool with a novel AHP based prioritization algorithm implemented, which can lead to discovery of new associations at SNP, gene, and pathway level. In near future, we expect that these new associations described through GWAS here and in other studies will lead to development of personalized medicine approaches with application in pharmacogenomics and psychopharmacology.