Figure 1.
Pipelines for calling single nucleotide variants (SNVs). SNVs were called in four sets, based on SAMtools: mpileup (SNP set1 and SNP set3) and GATK: unified genotyper (SNP set2 and SNP set4).
The numbers of reads and SNPs for individual steps are given for one exome-seq data set, generated using a Solexa GAIIx Genome Analyzer.
Table 1.
Validation rates according to TD and SNPQ categories in 4 analysis pipelines.
Figure 2.
Distribution of validation rate according to SNP quality (SNPQ) and total read depth (TD).
(A) Validation rate of SNPQ for SAMtools (SNP set1 and SNP set3). (B) Validation rate of SNPQ for GATK (SNP set2 and SNP set4). (C) Validation rate of TD for SNP set1–4.
Figure 3.
Diagram and validation rate of common variants.
(A) The diagram of common variants among the four types of SNP sets. (B) The validation rates of common variants and caller-specific variants.
Figure 4.
Evaluation of analysis parameters for improving validation rates.
(A) Distribution of validation rates according to genotype quality (GQ) values. (B) Distribution of validation rates according to mapping quality (MQ) values. (C) Distribution of validation rates according to alternate allele forward (AF) percent for SAMtools (SNP set1 and SNP set3). (D) Distribution of validation rates according to strand bias (SB) values for GATK (SNP set2 and SNP set4).
Figure 5.
Validation rates according to suggested cutoff values of parameters are shown in table (A) and graph (B) form using the SAMtools algorithm after realignment and recalibration.
A+B+C: filtered variants together with suggested values of SNPQ, GQ, and SB.
Figure 6.
Validation rates according to suggested cutoff values of parameters are shown in table (A) and graph (B) form using the GATK algorithm after realignment and recalibration.
A+B+C: filtered variants together with suggested values of SNPQ, GQ, and SB.