Prediction of Protein Functional Domains from Sequences Using Artificial Neural Networks

Table 2.

Comparison of Sequence Preprocessing Methods in Various Domain Groups

  Training set Test set Total
Group tp fp fn tn tp fp fn tn tp fp fn tn
EGF-like domain S 281 13 10 34988 134 13 11 17488 415 22 21 52480
R 291 5 0 329 145 4 0 136 436 7 0 467
Fibronectin III Domain S 214 1 23 35054 100 12 13 17521 325 11 25 52577
R 237 1 0 507 113 0 0 228 350 2 0 734
Sushi domain (SCR repeat) S 105 0 1 35186 59 5 0 17582 165 0 0 52938
R 106 0 0 77 59 0 0 32 165 0 0 109
ANK repeat S 98 0 4 35190 55 52 2 17539 151 3 6 52778
R 99 1 3 56 54 0 1 22 155 2 2 77
ABC transporters S 530 0 0 34762 239 0 0 17407 769 0 0 52169
R 530 3 0 90 239 2 0 43 769 1 0 137
WD repeat S 184 3 6 35099 77 40 0 17529 261 3 6 52668
R 186 2 4 168 76 0 1 92 262 2 5 260

  • S = whole sequence vs. R = regions.

This Article

  1. Genome Res. 11: 1410-1417

Preprint Server



Current Issue

  1. February 2026, 36 (2)
  1. Current Issue
  1. Alert me to new issues of Genome Research