Table 2.
Results for Rfam and small ncRNA benchmark set
| Quality (MERGED) |
Time (in s) |
||||||
|---|---|---|---|---|---|---|---|
| i | #Seq | #C | F | Rand | Phase 4 | Timei | TimeALL |
| Rfam benchmark | |||||||
| 0 | 8314 | 8314 | |||||
| 1 | 271 | 5 | 0.882 | 0.888 | 458 | 14 995 | 23 309 |
| 2 | 629 | 14 | 0.834 | 0.932 | 416 | 19 962 | 43 272 |
| 3 | 1076 | 23 | 0.868 | 0.956 | 334 | 15 108 | 58 380 |
| 7 | 2181 | 58 | 0.877 | 0.985 | 154 | 11 964 | 104 940 |
| 15 | 2821 | 130 | 0.834 | 0.984 | 77 | 2491 | 129 626 |
| Small ncRNA benchmark | |||||||
| 0 | 720 | 720 | |||||
| 1 | 140 | 10 | 0.942 | 0.945 | 42 | 2434 | 3154 |
| 2 | 232 | 20 | 0.926 | 0.939 | 27 | 3395 | 6549 |
| 3 | 270 | 26 | 0.936 | 0.935 | 17 | 7681 | 14 230 |
| 7 | 329 | 35 | 0.890 | 0.897 | 5 | 250 | 23 186 |
| 15 | 360 | 43 | 0.858 | 0.866 | 1 | 92 | 24 301 |
Results for each iteration i on the MERGED partition. Clustering quality is given as F measure and Rand index. The total number of clustered sequences is indicated with #Seq. The total number of clusters after merging is given by #C. Timei denotes the total time for iteration i, TimeALL is the total serial time up to iteration i.