Skip to main content
. 2024 Aug 7;40(8):btae493. doi: 10.1093/bioinformatics/btae493

Table 1.

Analysis of memory and runtime needed to produce the similarity matrixa

ModDotPlot StainedGlass
Sequence n (mbp) Plot Type w (bp) CPU time (s) Memory (GB) CPU time (s) Memory (GB)
Human CHM13
Chr1 Centromere
4.0 Static 4000 11.10 0.43 1871.31 12.95
Interactive 1000 204.85 1.16 2812.49 13.44
Gibbon mSymSyn1
ChrY
29.9 Static 32 000 51.16 2.05 9857.57 30.13
Interactive 8000 193.22 2.41 11 264.01 33.50
Human HG002
ChrY
62.5 Static 64 000 80.47 4.06 11 214.19 43.18
Interactive 16 000 269.84 5.90 14 806.91 48.95
ArabadopsisCol-CEN
Whole Genome
128.5
c = 5
Static 32 000 289.12 6.13 16 014.17 33.41
Interactive 8000 1734.11 9.57 20 187.19 35.20
Human CHM13
Whole Genome
3117.3
c = 24
Static 256 000 15 238.04 40.24
Interactive 64 000 29 101.76 44.31
a

This does not include plot runtime, as that is the same between StainedGlass and ModDotPlot. ModDotPlot was run with a target sketch size of m =1000 for all samples. For the whole genome assemblies of Arabadopsis and CHM13, the runtime includes the comparative matrix between each pairwise combination of chromosomes, in addition to self-identity comparisons. StainedGlass was unable to complete CHM13 whole genome within 72 h of CPU time.