Table 1.
Accuracy of models before and after refinement using high-resolution cryo-EM maps and hybrid density maps of 5, 6, and 8 Å resolutions
| Case a | #Amino acids b | Res. (Å) c | AF2 pLDDT d | AF2TM e | Refined models f | |||
|---|---|---|---|---|---|---|---|---|
| High | 5 Å | 6 Å | 8 Å | |||||
| 7LV9-23530-B | 97 | 4.5 | 55.37 | 0.39 | 0.42 | 0.23 | 0.34 | 0.34 |
| T1047S1D1-7BGL-12183-A | 211 | 2.2 | 75.74 | 0.53 | 0.71 | 0.50 | 0.51 | 0.49 g |
| T1047S2D3-7BGL-12183-A | 116 | 2.2 | 75.17 | 0.71 | 0.76 | 0.51 g | 0.55 g | 0.51 g |
| 7KZZ-23093-B | 281 | 3.42 | 89.76 | 0.76 | 0.81 | 0.80 | 0.70 | 0.66 |
| 7KU7-23035-A | 269 | 3.4 | 88.56 | 0.76 | 0.79 | 0.79 | 0.78 | 0.71 |
| 7LCI-23274-R | 393 | 2.9 | 81.15 | 0.82 | 0.95 | 0.94 | 0.94 | 0.64 |
| 7M7B-23709-A | 209 | 2.95 | 77.60 | 0.83 | 0.91 | 0.81 | 0.81 | 0.81 |
| 7MLZ-23914-A | 196 | 3.71 | 83.09 | 0.84 | 0.90 | 0.88 | 0.87 | 0.61 |
| 7MJS-23883-H | 132 | 3.03 | 89.57 | 0.88 | 0.98 | 0.92 | 0.88 | 0.63 g |
| 7BRM-30160-A | 257 | 3.6 | 87.52 | 0.89 | 0.89 | 0.94 | 0.90 g | 0.84 |
| 7LX5-23566-B | 196 | 3.44 | 82.93 | 0.89 | 0.93 | 0.93 | 0.92 | 0.64 |
| 7EDA-31062-A | 334 | 2.78 | 89.02 | 0.90 | 0.92 | 0.93 | 0.86 | 0.66 g |
| 7L6U-23208-A | 311 | 3.3 | 89.74 | 0.90 | 0.96 | 0.95 | 0.95 | 0.93 |
Protein IDs (PDB ID_EMDB ID_Chain ID). For the two chains involved in CASP challenges, CASP target IDs are indicated.
The number of amino acids in the protein.
The resolution of cryo-EM maps.
The average pLDDT scores of AlphaFold2 models.
The accuracy is indicated as TM scores for models obtained from AlphaFold2.
TM scores for models refined using Phenix (Terwilliger et al., 2022), the cryo-EM maps (High) and the hybrid density maps at 5, 6, and 8 Å, respectively.
The Phenix resolution parameter was tuned 2–3 Å lower than the nominal resolution of the density map to ensure completion of the refinement protocol (see text).