Fig. 5

Improved assembly using updated methods. Distribution of fragment sizes and complexity for contigs assembled using old (MinKNOW/Canu v1.4; a) and new (Albacore/Canu v1.5; b) methods with unamplified DNA. The structures of the assembled contigs (drawn to scale, as indicated) are also shown. The unbranched subgraphs representing more than 70% of the sequence (dark magenta bars) have been omitted. The distribution of fragment sizes from the WTSI reference genome (dashed dotted line) is shown in (b). The largest two subgraphs in (a) contain 207 contigs (total of 28.7 Mb) and 101 contigs (17.3 Mb) and in (b) 106 contigs (13.3 Mb) and 76 contigs (10.6 Mb). In addition to the better resolution of subgraphs, in the new assembly, there is a marked increase in unbranched subgraphs longer than 1 Mb. This improvement in contiguity reflects the increase in the quality and quantity of called reads, and better assembly algorithms. WTSI Wellcome Trust Sanger Institute