Table 2.

Comparison of de novo assembly quality using TransRate contig and read‐mapping metrics. Bridger, SOAPdenovo‐Trans, and Trans‐ABySS assemblies are multi‐k‐mer assemblies, see Methods for details. See Smith‐Unna et al. (2016) and http://hibberdlab.com/transrate/metrics.html for further details regarding TransRate metrics. Read‐mapping metrics were not considered appropriate for the coding assembly, which is a selective dataset

Metric	Bridger	SOAPdenovo‐Trans	Trans‐ABySS	Trinity	Total assembly	Coding assembly
No. of contigs	513,499	282,833	318,436	157,274	484,080	147,450
Shortest	189	200	200	201	200	297
Longest	27,543	17,265	37,689	11,148	34,468	34,468
Mean length	780	445	573	503	726	1,121
No. with ORF	138,937	51,336	72,597	31,566	135,969	100,823
ORFs %	27%	18%	23%	20%	28%	81%
N50	1,218	480	738	602	1,071	1,518
% of reads mapped	97%	93%	92%	82%	98%	–
Good mappings	30,880,238	28,292,720	28,688,483	22,012,681	31,833,986	–
% good mappings	88%	81%	82%	63%	91%	–
Potential bridges	43,048	46,708	38,398	38,011	34,120	–
% uncovered bases	38%	11%	15%	4%	23%	–
% contigs with uncovered bases	86%	71%	77%	44%	80%	–
% contigs uncovered	45%	16%	18%	4%	29%	–
% contigs low‐covered	93%	75%	80%	76%	89%	–
Transrate score	0.13	0.26	0.25	0.21	0.42	–