Differentiating Genomic SNPs Using Allele Depth and Predicted Genotype

Abstract

Differentiating between genomic SNPs and other types of single nucleotide variants becomes a key issue in research aimed at studying the importance of these variants of a particular type in biological processes. Here we present an R based method for differentiating between genomic single nucleotide polymorphisms (SNPs) and RNA editing sites. We use data from an earlier study of ours and target only the known dbsnp SNPs that we found in our study. Our method involves calculating the ratio of allele depth for ref and alt alleles and comparing that to the predicted genotype. We use the concept that editing levels should be different for each allele and thus should not reflect the ratio predicted by the genotype. The study yielded an accuracy rate ranging from 86 to over 90 percent at successfully predicted dbsnp entries as SNPs. Albeit this is in the absence of known RNA editing site vcf data to compare as a reference.