Skip to main content
. 2018 May 9;2018:1391265. doi: 10.1155/2018/1391265

Table 8.

The number of host proteins shared by training (TR) and test (TS) datasets used for assessing the applicability of the SVM model to new viruses and to new hosts.

Dataset TR1 TS1 TR1 TS2 TR1 TS3 TR1 TS4 TR1 TS5
#PPIs 638 515 638 30 638 377 638 319 638 1578
#Virus proteins 25 11 25 12 25 10 25 11 25 46
#Host proteins 499 424 499 27 499 307 499 298 499 1056
#Host proteins common to TR and TS 63 (14.9%) 5 (18.5%) 68 (22.1%) 22 (7.4%) 122 (11.6%)

Dataset TR2 TS6 TR2 TS7 TR2 TS8 TR2 TS9 TR2 TS10
#PPIs 689 191 689 125 689 86 689 57 689 78
#Virus proteins 35 116 35 34 35 24 35 10 35 27
#Host proteins 522 141 522 87 522 79 522 38 522 64
#Virus proteins common to TR and TS 9 (7.8%) 1 (2.9%) 4 (16.7%) 0 (0.0%) 0 (0.0%)

The numbers in parentheses represent the proportion of common proteins to proteins in test datasets.