Figure 3

An example of alignments near a false-positive SNP position with no quality control or with trimming. (a) no quality control – BWA, (b) trimming – BWA, (c) masking – BWA, (d) no quality control – BOWTIE2, (e) trimming – BOWTIE2, and (f) masking – BOWTIE2. The top, the second, and the third line in an alignment picture show a base pair position, the reference genome sequence, and the consensus sequence, for each. A base call that matches the reference sequence is denoted as a dot for the forward strand or as a comma for the reverse strand. The colors in a read denote the base call quality. White is for a base with its error rate smaller than 0.1%, yellow for 0.1 ~ 1% error rate, green for 1 ~ 10% error rate, and blue for 10% or greater error rate. Since we used 1% cutoff, all the green and the blue were either trimmed off or masked as ‘N’ or ‘n’. The graphics were made using tview in Samtools (version 0.1.18) [19]. BOWTIE2 was used to align reads. GATK haplotype SNP caller made false-positive homozygous SNP calling in (a), (b), (d), and (e), not in (c) and (f).