Skip to main content
. 2008 Sep 25;24(23):2780–2781. doi: 10.1093/bioinformatics/btn507

Table 1.

Comparison of DaliLite v.3, the SSM server and SCOP

Query Dali −q time (min) Dali −q P Dali −q Z AUC Dali −l time (min) Dali −l Z AUC SSM Q AUC SCOP T SCOP class
1c52 14 241 0.885 297 0.822 0.674 219 a.3.1.
1sfxA 12 401 0.757 357 0.733 0.386 331 a.4.5.
1azu 14 529 0.967 369 0.969 0.516 289 b.6.1.
1wk2A 7 49 0.800 201 0.966 0.197 60 b.122.1
3be7A 51 542 0.990 1861 0.968 0.843 135 c.1.9.
1qlwA 30 414 0.908 1211 0.897 0.576 456 c.69.1.
2baa 18 726 0.693 486 0.354 0.025 979 d.2.1
1wotA 13 500 0.558 251 0.642 0.009 179 d.218.1

AUC, area under the curve of reliability (TP/P) versus coverage (TP/T), where Ts (‘true’) are members of the same SCOP (Murzin et al.1995) superfamily as the query structure, Ps (‘positive’) are the top n matches from the ranked list for varying n and TPs (‘true positive’) are the intersection of sets T and P. Only PDB entries classified in SCOP v.1.73 were evaluated and each PDB entry was counted once. SSM server http://www.ebi.ac.uk/msd-srv/ssm/cgi-bin/ssmserver; (Krissinel and Henrick, 2004) parameters were set to 10 : 10 and highest precision. The DaliLite search was performed using the –list (Dali –l) or –quick option (Dali –q with MAX_HITS=1000) and reporting matches with Z>2.