Skip to main content

Table 2 Comparison of contigs produced from in silico reads from the BASV virus

From: ContigExtender: a new approach to improving de novo sequence assembly for viral metagenomics data

BASV (11.9 kb) Error rate = 0.01 Error rate = 0.05
Read length (bp) Depth (x) Spike-in peaks Contig Extender (kb) PRICE (kb) GenSeed (kb) Kollector (kb) Meta SPAdes (kb) Contig Extender (kb) PRICE (kb) GenSeed (kb) Kollector (kb) Meta SPAdes (kb)
100 10   6.7 1.6 1.1 1.8 6.7 1.9 1.3 NA NA 1.8
  20   11.9 11.9 1.1 2.7 11.9 11.9 11.8 NA NA 7.9
  20 Yes 11.9 11.8 1.1 2.7 11.9 11.9 11.9 NA NA 7.9
  50   11.9 11.9 1.1 4.9 11.9 11.9 11.8 1.1 NA 7.9
250 10   11.8 7.2 1.1 1.1 7.9 8.4 7.3 NA NA 7.6
  20   11.9 11.7 1.3 2.3 7.9 11.8 11.8 NA NA 7.8
  20 Yes 11.9 11.5 1.3 2.3 7.9 11.8 NA NA NA 7.9
  50   11.9 11.8 1.4 2.4 7.9 11.9 11.9 NA NA 11.9
  1. Longest contig length produced by ContigExtender and other tools using in silico synthetic sequences for the BASV virus, representing varying read lengths, error rates, depths, and unevenness of coverage. Randomly selected sequences of 1 kb were used as seed contigs. Spike-in sequences were added, with each peak at a depth of at 50 × coverage and spanning 250 bp. De novo assemblies using metaSPAdes were also performed as a benchmark for assembly difficulty for each dataset. Runs that fail to produce extension are marked “NA”