Skip to main content

Table 1 Results across all users for BioCreAtIvE Task 2.1.

From: Protein annotation as term categorization in the gene ontology using word proximity networks

User, Run # results "perfect" "generally"
4, 1 1048 268 (25.57%) 74 (7.06%)
5, 1 1053 166 (15.76%) 77 (7.31%)
5, 2 1050 166 (15.81%) 90 (8.57%)
5, 3 1050 154 (14.67%) 86 (8.19%)
7, 1 1050 263 (25.05%) 150 (14.29%)
7, 2 1856 43 (2.32%) 40 (2.16%)
7, 3 1698 59 (3.47%) 27 (1.59%)
9, 1 251 125 (49.80%) 13 (5.18%)
9, 2 70 33 (47.14%) 5 (7.14%)
9, 3 89 41 (46.07%) 7 (7.87%)
10, 1 45 36 (80.00%) 3 (6.67%)
10, 2 59 45 (76.27%) 2 (3.39%)
10, 3 64 50 (78.12%) 4 (6.25%)
14, 1 1050 303 (28.86%) 69 (6.57%)
15, 1 524 59 (11.26%) 28 (5.34%)
15, 2 998 125 (12.53%) 69 (6.91%)
17, 1 412 0 (0.00%) 1 (0.24%)
17, 2 458 1 (0.22%) 0 (0.00%)
20, 1 1048 300 (28.63%) 57 (5.44%)
20, 2 1050 280 (26.72%) 60 (5.73%)
20, 3 1050 239 (22.76%) 59 (5.62%)
  1. Evaluation results on the evidence text selected for Task 2.1. A "perfect" evaluation indicates that the evidence text refers to both the correct protein and the correct GO node. A "generally" evaluation indicates that it refers to the correct protein and that the reference to a GO node is somewhat too general. The Los Alamos team is user 7.