Table 4 Model selection ability on ab initio datasets for DeepQA, ProQ2, Dope2, and RWplus score based on TM-score and RMSD, and their summation of Z-score
From: DeepQA: improving the estimation of single protein model quality with deep belief networks
QA methods | TM-score on top 1 model/SUM Z-score (>0.0) | RMSD on top 1 model/SUM Z-score (<0.0) | TM-score on best of top 5/SUM Z-score (>0.0) | RMSD on best of top 5/SUM Z-score (<0.0) |
|---|---|---|---|---|
DeepQA | 0.23/0.86 | 19.01/-0.76 | 0.26/1.78 | 17.14/-1.52 |
ProQ2 | 0.22/0.40 | 19.73/-0.37 | 0.25/1.28 | 17.93/-1.04 |
Dope | 0.22/0.49 | 19.55/-0.51 | 0.24/1.13 | 18.10/-1.00 |
RWplus | 0.22/0.53 | 19.68/-0.35 | 0.25/1.49 | 17.38/-1.41 |
Mean | 0.22/0.68 | 19.49/-0.64 | 0.25/1.46 | 17.64/-1.26 |