Skip to main content

Table 1 Comparison between F measure while using Blocks and k-length subsequences

From: A method for probabilistic mapping between protein structure and function taxonomies through cross training

Class

F (subsequences)

F (Blocks)

Small proteins

.78

.73

Globin like

.71

.684

NAD P binding Rossmann fold domains

.54

.537

  1. Comparison of F-measure using k-length overlapping subsequences of length (k) equal to 4 and blocks as features. In the case of using blocks as features, F-measure was found to be only slightly lower for all the classes, while the size of feature set was many times smaller.