Fig. 2

Distribution of read length. A 5.3 kb-long insert library was sequenced at 180 min. The raw reads were filtered and split based on adapters to generate sub-reads referred to here as ‘reads’. The length of each read is plotted against the read counts. The first peak corresponds to reads with length around 2.7 kb and the second peak corresponds to reads that spanned the full-length 5 kb amplicon. Most of the reads were shorter than the amplicon size, but there was a substantial number of long reads that spanned the full-length amplicon. With the very long reads (>5 kb), the adapters delimiting the subreads were either missing or not identified