Skip to main content

Table 2 Effect of motif length on clustering performance.

From: Clustering of protein domains for functional and evolutionary studies

  motif length  
  5 10 15 20 25 30  
Protein family false assignments No. sequences
Nucleotidyl cyclases 0 0 0 0 0 0 75
Protein kinases 0 0 0 0 0 0 215
MDH/LDH 5 6 5 5 4 4 183
AT-domains 2 3 4 4 5 5 181
KR-domains 20 18 20 17 10 9 72
sHSP 10 13 14 11 5 5 214
  1. The amino acids positions with the highest evolutionary split scores were used to construct the motifs.