Detecting SNPs and estimating allele frequencies in clonal bacterial populations by sequencing pooled DNA

Summary: Here, we present a method for estimating the frequencies of SNP alleles present within pooled samples of DNA using high-throughput short-read sequencing. The method was tested on real data from six strains of the highly monomorphic pathogen Salmonella Paratyphi A, sequenced individually and in a pool. A variety of read mapping and quality-weighting procedures were tested to determine the optimal parameters, which afforded [≥]80% sensitivity of SNP detection and strong correlation with true SNP frequency at poolwide read depth of 40x, declining only slightly at read depths 20-40x. Availability: The method was implemented in Perl and relies on the opensource software Maq for read mapping and SNP calling. The Perl script is freely available from ftp://ftp.sanger.ac.uk/pub/pathogens/pools/. Contact: kh2@sanger.ac.uk Supplementary information: Supplementary data are available at Bioinformatics online. 10.1093/bioinformatics/btp344