Table 1.
AlphaFold2 predictive modeling on the cryoEM challenge data sets
| EMDB/PDB ID (resolution) | Length (aa) | Molprobity score | Clash score | Ramachandran outliers (%) | Rotamer outliers (%) | Map/model CC | pLDDT | TM-score | RMSD (Å) |
|---|---|---|---|---|---|---|---|---|---|
| EMD 0406/6NBB (2.9 Å) | 375 | 1.88 | 15.03 | 0.54 | 0.97 | 0.7825 | 96.77 | 0.7754 | 3.790 (374) |
| EMD 2677/5A63 (4.5 Å) | |||||||||
| A | 709 | 2.22 | 19.81 | 2.40 | 0.65 | 0.8013 | 90.96 | 0.9795 | 1.581 (665) |
| B | 467 | 3.56 | 52.33 | 16.99 | 5.88 | 0.5760 | 66.51 | 0.9472 | 1.420 (215) |
| C | 265 | 1.87 | 17.59 | 1.14 | 0.93 | 0.7714 | 93.06 | 0.9753 | 0.985 (243) |
| D | 101 | 1.94 | 25.78 | 0.00 | 1.12 | 0.8004 | 93.12 | 0.9417 | 1.059 (100) |
| EMD 2842/4UDV (3.3 Å) | 159 | 2.47 | 15.83 | 3.82 | 2.16 | 0.6328 | 88.55 | 0.5901 | 3.935 (153) |
| EMD 2847/5AFI (2.9 Å) | 201 | 1.77 | 18.94 | 0.50 | 0.0 | 0.5792 | 93.41 | 0.9420 | 1.364 (201) |
| E | 177 | 1.80 | 20.28 | 0.0 | 0.72 | 0.7169 | 91.34 | 0.6339 | 3.801 (176) |
| G | 142 | 1.91 | 24.46 | 0.0 | 0.86 | 0.7076 | 93.62 | 0.9687 | 0.945 (142) |
| J d< | 206 | 1.71 | 16.03 | 0.0 | 0.0 | 0.6944 | 92.32 | 0.6757 | 3.752 (205) |
| g | 179 | 2.68 | 16.06 | 6.78 | 3.40 | 0.6235 | 92.16 | 0.6108 | 3.852 (151) |
| o | 89 | 1.48 | 8.87 | 0.0 | 0.0 | 0.687 | 92.80 | 0.4674 | 3.818 (88) |
| EMD 5623/3J9I (3.3 Å) | 233 | 1.62 | 12.83 | 0.43 | 0.0 | 0.8474 | 94.32 | 0.9694 | 1.076 (224) |
| EMD 5778/3J5P (3.275 Å) | 838 | 3.20 | 30.04 | 14.71 | 3.91 | 0.6057 | 69.99 | 0.9545 | 2.963 (592) |
| EMD 6000/3J7N (3.8 Å) | 189 | 3.39 | 60.10 | 11.76 | 3.42 | 0.6241 | 66.90 | 0.7805 | 3.649 (149) |
| EMD 6422/3CAU (4.1 Å) | 548 | 2.06 | 15.28 | 2.01 | 1.20 | 0.6460 | 92.21 | 0.7346 | 5.354 (526) |
| EMD 7770/6CVM (1.9 Å) | 1,021 | 1.97 | 22.77 | 0.29 | 0.34 | 0.7100 | 94.45 | 0.8837 | 3.850 (1020) |
| EMD 20027/3AJO (2.32 Å) | 183 | 1.86 | 13.73 | 2.21 | 0.61 | 0.7675 | 95.10 | 0.6305 | 3.785 (172) |
| EMD 22898/3KJR (2.08 Å) | |||||||||
| A | 284 | 3.62 | 30.09 | 18.79 | 12.84 | 0.2329 | 48.16 | 0.4580 | 11.507 (193) |
| AB∗ | 568 | 3.58 | 27.67 | 15.96 | 13.04 | 0.2756 | 39.11 | 0.6035 | 9.469 (193) |
| C | 210 | 2.57 | 5.53 | 4.81 | 15.68 | 0.2674 | 72.43 | 0.1247 | 2.756 (31) |
| EMD 30210/7KVB (2.5 Å) | |||||||||
| A | 951 | 2.65 | 35.57 | 3.48 | 1.55 | 0.5375 | 88.85 | 0.8402 | 4.675 (834) |
| B | 207 | 2.26 | 17.64 | 0.49 | 1.70 | 0.4233 | 85.04 | 0.4654 | 5.535 (114) |
| C | 92 | 2.90 | 33.45 | 3.33 | 3.49 | 0.5366 | 80.14 | 0.3089 | 3.899 (63) |
AlphaFold2 was run on each protein subunit from the individual cryoEM challenge data sets, with the exception of EMDB: EMD-2847. For EMDB: EMD-2847, a 70s ribosome structure with over 50 individual protein subunits, six random chains were selected for modeling with AlphaFold2. The resulting models were assessed for model quality including the Molprobity score, clash score, Ramachandran outliers, rotamer outliers), fit to density (map/model cross correlation), predicted level of confidence (pLDDT), and similarity to the reference model (TM-score, RMSD). In the final column, the number listed in parenthesis after RMSD refers to the total number of matched residues between the model and reference structures. The resolution for each map is listed in the first column after the EMDB and PDB IDs. For EMDB: EMD-22898, the asterisk indicates AlphaFold2 was run on a homodimer of chain A, denoted as AB in the table and text.