Table 2.
Data Set | BLASTClust | BAG | SEQOPTICS | ||||||
S | P | R | S | P | R | S | P | R | |
l(Pfam) | 0.05 | 1.00 | 0.05 | 0.27 | 1.00 | 0.27 | 0.83 | 0.99 | 0.84 |
2(Pfam) | 0.04 | 1.00 | 0.04 | 0.20 | 1.00 | 0.20 | 0.85 | 0.98 | 0.87 |
3(NCBI) | 0.11 | 1.00 | 0.11 | 0.60 | 1.00 | 0.60 | 0.66 | 0.82 | 0.78 |
4(Swiss-Prot) | 0.06 | 1.00 | 0.06 | 0.50 | 1.00 | 0.50 | 0.81 | 0.99 | 0.82 |
Clustering results of four data sets by three methods according to three parameters: S(Jaccard Coefficient), P(Precision), R(Recall). It shows that SEQOPTICS outperforms BAG and BLASTClust which tend to give more clusters than "true" clusters.