Sodium bisulfite conversion followed by sequencing (BS-Seq, such as whole genome bisulfite sequencing or reduced representation bisulfite sequencing) has become popular for studying human epigenetic profiles. Identifying single nucleotide polymorphisms (SNPs) is important for quantification of methylation levels and for study of allele-specific epigenetic events such as imprinting. However, SNP calling in such data is complex and time consuming. Here we present an ultrafast and memory-efficient package named BS-SNPer for the exploration of SNP sites from BS-Seq data. Compared to Bis-SNP, a popular BS-Seq specific SNP caller, BS-SNPer is over 100 times faster and uses less memory. BS-SNPer also offers higher sensitivity and specificity compared to existing methods.

AVAILABILITY AND IMPLEMENTATION: BS-SNPer is written in C++ and Perl, and is freely available at https://github.com/hellbelly/BS-Snper.

CONTACT: bolund@biomed.au.dk, kdso@clin.au.dk or orntoft@ki.au.dk SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.