Data integration for functional annotation of regulatory single nucleotide polymorphisms associated with Alzheimer's disease susceptibility

Abstract

Background: Alzheimer's disease (AD), the most common form of dementia affects 24.3 million people worldwide. More than twenty genetic loci have been associated with AD and a significant number of genetic variants were mapped within these loci. A large proportion of genome wide significant variants lie outside the coding region. However, the plausible function of these variants is still unexplored.

Objective: The present study aimed to unravel the regulatory role of proxy single nucleotide polymorphisms (SNPs), to determine their risk of developing AD.

Methods: The RegulomeDB was employed to predict the regulatory role of proxy SNPs. Protein association network and functional enrichment analysis was performed using String10.5 and gene ontology, respectively.

Results: A total of 451 SNPs were examined through SNAP web portal (r² ≤ 0.80) which returned 2186 proxy SNPs in linkage disequilibrium (LD) with genome wide significant SNPs for AD. Out of 2186 SNPs analyzed in RegulomeDB, 151 had the scores < 3 that indicates the high degree of their potential regulatory function. Further analysis revealed that out of these 151 SNPs, 37 were genome wide significant for AD, 17 were significantly associated with diseases other than AD, 89 were proxy SNPs (not genome wide significant) for various diseases including AD while 8 SNPs were novel proxy SNPs for AD.

Conclusion: These findings support the notion that the non-coding variants can be strongly associated with disease risk. Further validation through genome wide association studies will be helpful for the elucidation of their regulatory potential.

Keywords: Alzheimer's disease; Genome wide association studies; Linkage disequilibrium; Non-coding variants; RegulomeDB; Single nucleotide polymorphisms.