Briefly, while the details depend on the technology, these plots document changes in error rates at GC extremes and on long homopolymers, for every technology. For example, Illumina, which had very low insertion and deletion error rates, had a substantial rise in insertions and deletion rates at both GC extremes. The Ion Torrent insertion and deletion rates were more consistent, albeit higher than Illumina's, across a range of GC contents, but the mismatch rate was elevated at low and high-GC regions. As another example, we note that for Pacific Biosciences, the deletion rate rose at high GC, while the insertion rate declined. This behavior appears to result from lower signal-to-noise ratios for the dyes attached to G and C bases in C1 chemistry (personal communication, Edwin Hauw, Pacific Biosciences, USA). Complete Genomics showed consistent (relatively high) mismatch and (relatively low) insertion rates across the GC spectrum, but the deletion rate rose substantially at the extremes. Within long homopolymers, the behavior of insertion and deletion errors would depend on whether a technology systematically over- or under-reports homopolymer length. For example, as
