ARACHNE: A Whole-Genome Shotgun Assembler

Table 5.

Misassemblies

Genome H. influenzae S. cerevisiae D. melanogaster Human 21 Human 22
Full Coverage (∼10-fold)
 Deletions 2 5 102 13 26
 Mean length (bp) 440 470 1660 360 430
 Insertions 1 7 2
 Mean length (bp) 350 990 400
 Hanging ends 3 4
 Mean length (bp) 190 800
 Other 3 1
 Misassemblies (note) (note)
Half Coverage (∼5-fold)
 Deletions 3 4 116 26 41
 Mean length (bp) 290 3790 1600 220 340
 Insertions 2 12 2 3
 Mean length (bp) 380 670 90 390
 Hanging ends 1 2 42 11 14
 Mean length (bp) 78 450 460 1330 826
 Other 5 4 5
 misassemblies (note) (note) (note)
  • See Figure 6A,B for an illustration of the various types of common misassemblies. Other misassemblies are as follows:

  • Two contigs of lengths 1 kb and 138 kb were exchanged in a supercontig. Two correct contigs of lengths 313 kb and 169 kb, which were separated in the genome by 60 bp, were glued along a 440-bp segment that appeared at the left end of the left contig, and also at the right end of the right contig, yielding a chimeric contig in the final assembly (which we call a slipped join). There was a standard misassembly (Fig. 6B), occurring in a 4.8-Mb supercontig.

  • At the end of a supercontig, there was a contig of length 10 kb, which aligned at a distant location in the genome relative to the rest of the supercontig.

  • A supercontig of 467 kb had a standard major misassembly. A supercontig of 1.4 Mb had two stray contigs of 5 kb and 2 kb. A contig had a deletion of 29 kb. Contigs of 27 kb and 37 kb were misassembled.

  • A 3.4-Mb supercontig had a standard major misassembly. Contigs of 9 kb and 24 kb had slipped joins. A 12-kb contig at the end of a 2.6-Mb supercontig was misassembled.

  • A 4.5-Mb supercontig had a standard major misassembly. A 14-kb contig had a slipped join. A 303-kb supercontig was slightly misassembled near one of its ends. A supercontig had a spurious contig of length 5 kb. A 9-kb supercontig consisted of two incorrectly linked contigs.

This Article

  1. Genome Res. 12: 177-189

Preprint Server