SV-AUTOPILOT: optimized, automated construction of structural variation discovery and benchmarking pipelines

BMC Genomics

Table 3 Typical computational performance by SV tools used for a single run

Tool	Multi- threading	Mem use on Tair (Mb)	Mem use on Human (Mb)	CPU time Tair (h:m.s)	CPU time Human (h:m.s)	Algorithm	SV’s
GASV	n	1058	594	0:02.08	0:01.20	PE	IDVT
Delly	n	578	1236	0:15.02	0:03.18	PE & SR	DVTP
Breakdancer	n	21.9	7	0:02.41	0:27.7	PE	IDVT
Pindel	y	3500.5	5779	3:02.46	1:16.0	SR	IDVP
Clever	y	238.7	1598	0:15.47	0:14.04	PE	ID
SVdetect	n	172.3	3223	0:07.56	0:07.31	PE	IDVTP
Prism	n	1024.9	6817	0:28.15	0:05.59	PE & SR	IDVP

Log files document computation performance for each tool used in this benchmarking study. Documentation from a single run shows memory (mem) usage and CPU time need to run each tool on Arabidopsis (Tair) and on the Human dataset used in the benchmarking. Additional columns refer to the type of algorithm used (PE: Paired-end; SR: Split-read) and the SVs that the tool is reported to be able to predict (I: Insertion; D: Deletion; V: Inversion; T: Translocation; P: Duplication). Raw log files are included in the supplementary data.

ISSN: 1471-2164