Method | Enriched Clusters | # NEC | % NEC | # NEC S | % NEC S |
---|
PPI | 234/324 (72.22%) | 3082.0 | 50.54% | 2158.0 | 35.39% |
4.5 | 194/266 (72.93%) | 1647.0 | 27.02% | 1330.0 | 21.82% |
5.0 | 199/309 (64.40%) | 3589.0 | 58.87% | 2203.0 | 36.14% |
5.5 | 189/291 (64.95%) | 3765.0 | 61.76% | 2228.0 | 36.55% |
6.0 | 177/249 (71.08%) | 4670.0 |
76.61%
| 2490.0 |
40.85%
|
- NEC= “Nodes in Enriched Clusters”. We calculate %NEC in two settings: %NEC is enrichment in the GO hierarchy with terms above the fifth level filtered out, and %NEC S uses the same filtered GO hierarchy, but then only gives a node credit if there is a match between one of the node’s labels and one of the terms for which there is GO enrichment for the cluster. For every threshold we tested ≥ 5, the percentage of nodes in enriched clusters is better than Spectral run alone for both measures
- Bolded values represent the best values achieved for the %NEC and %NEC S statistics comparing the PPI network and different DSD detangling thresholds