HaplotypeTools: a toolkit for accurately identifying recombination and recombinant genotypes

BMC Bioinformatics

Table 1 Accuracy of heterozygous variant calling by Pilon was assessed

Paired reads (100nt or 10 kb) were simulated at 20X depth from reference Bd JEL423 genome that was duplicated to create an in silico diploid. In silico mutations were then randomly introduced throughout (1/kb, 10/kb or 100/kb). Reads were aligned to the original reference sequence (non-duplicated, non-mutated version), and diploid variants called by Pilon. Counts of variants are shown including single nucleotide polymorphisms (SNP), heterozygous positions (HET), insertions (INS), deletions (DEL) and ambiguous (AMB). Accuracy was assessed according to Comparison of FDR tool [28], that calculated TN = true negatives (correct reference bases), TP = true positives (correct HET), FN = false negatives (incorrect reference bases) and FP = false positives (incorrect HET). FP (other) is a count of all additional (non-heterozygous) incorrect bases including SNPs, INS, DEL and AMB. > 99% of FP (other) were SNPs. TP (%) and FN (%) are precents of Introduced HET, FN (%) is a percent of assembly length, and FP (%) is a percent of HETs called. Sensitivity = TP/(TP + FN), Specificity = TN/(TN + FP + FP (other)), Accuracy = (TN + TP)/(TN + TP + FN + FP + FP (other))

ISSN: 1471-2105