Skip to main content
. Author manuscript; available in PMC: 2008 Aug 18.
Published in final edited form as: Pac Symp Biocomput. 2008:640–651.

Table 1.

Six automated methods for identifying mutant PDB entries are assessed against (a) manually curated mutant entry data, and (b) manually deposited mutation annotations. True positives (TP), false positives (FP), false negatives (FN), precision (P), recall (R), and F-meaure (F) are presented. Sequence validation (methods 2 and 4) is described in Section 2.1.3

(a) Mutant Entry Id. TP FP FN P R F
1 MutationFinder 2690 1499 5263 0.642 0.338 0.443
2 MutationFinder + validation 1665 396 6288 0.808 0.209 0.333
3 Alignment 6079 3728 1874 0.620 0.764 0.685
4 Alignment + validation 4104 2549 3849 0.617 0.516 0.562
5 2 and 3 1403 275 6550 0.836 0.176 0.291
6 2 or 3 6258 3816 1695 0.621 0.787 0.694
(b) Normalized Mutations TP FP FN P R F
1 MutationFinder 2575 7050 7929 0.268 0.245 0.256
2 MutationFinder + validation 1803 799 8701 0.693 0.172 0.275
3 Alignment 7681 15404 2823 0.333 0.731 0.457
4 Alignment + validation 5059 9225 5455 0.354 0.482 0.408
5 2 and 3 1584 532 8920 0.749 0.151 0.251
6 2 or 3 7900 15671 2604 0.335 0.752 0.464