Skip to main content

Table 6 Feature space sizes across sources

From: Learning virulent proteins from integrated query networks

Data src.

Num. features

Feature type

AmiGO

5102

terms

BioCyc

1674

proteins, pathways

Cdd

6463

models

GenNav

6425

terms

InterPro

3540

models

Kegg

234

pathways

Pdb

7954

structures

TigrFam

1109

models

  1. Number of features per source used for specific virulence predictions. Individual source feature sizes are reported before any feature selection.