Table 2.
Evaluation of the performance of individual EpiPredictor components against three genome-wide ChIP studies in D. melanogaster and their intersection
| Number of top genes | EpiPredictor Components | Schwartz et al.a | Tolhuis et al.b | Schuettengruber et al.c | Intersectiond |
|---|---|---|---|---|---|
| 243 e | (a) | 14.20%f | 5.33% | 12.09% | 2.63% |
| (a,b) | 22.73% | 9.78% | 19.53% | 23.68% | |
| (a,b,c) | 26.14% | 10.22% | 25.12% | 23.68% | |
| (a,b,c,d)g | 27.27% | 10.67% | 26.05% | 26.32% | |
| 322 h | (a,b,c,d) | 32.39% | 14.22% | 30.70% | 34.21% |
| (a,b,c,d,e)i | 35.80% | 15.11% | 33.02% | 44.74% |
(a): Motif Analyzer; (b): SVM Classifier; (c): GC Analyzer; (d): Conservation Level Analyzer; (e): Comparative Genomics Analyzer.
aOverlap with the genes predicted by Schwartz et al. (4).
bOverlap with the genes predicted by Tolhuis et al. (15).
cOverlap with the genes predicted by Schuettengruber et al. (14).
dOverlap with the genes intersected by Schwartz et al., Tolhuis et al., and Schuettengruber et al.
eThe number of top genes retrieved from EpiPredictor-Basic analysis.
fSuppose the validation set includes V genes. Among the top N genes predicted by our system, W genes matched the validation set, the overlap was represented as W/V.
gThe EpiPredictor-Basic module.
hThe number of top genes retrieved from EpiPredictor-CG analysis.
iThe EpiPredictor-CG module. The results corresponding to the EpiPredictor-Basic and EpiPredictor-CG models are highlighted in bold.