Figure 7From: Estimation of sequencing error rates in short readsTwo PhiX samples run on Illumina Genome Analyzer II: Proportion of reads containing mismatches by starting position. For a few positions, almost 100% of the reads covering them do not match the reference genome, substantially inflating the error rate estimates given by mismatch-counting. The widths of the bars reaching almost 1 (100%) and containing 1 mismatch are 36 bases, which is the read length for the PhiX samples.Back to article page