Figure 3
Figure 3 . AlphaFold2 models of AR9 nvRNAP proteins fit the cryo-EM density nearly perfectly. The cryo-EM-derived structures of gp105, gp154, and two gp226 domains are colored according to the color code given in the upper left corner of each panel. All AlphaFold2 models are colored magenta. The electron density is contoured at 4.25 standard deviations above the mean and colored semi-transparent grey. Regions where no cryo-EM-derived structure existed prior to the availability of the AlphaFold2 models are indicated with a dashed line and their boundary residues are labeled.
The accuracy of AlphaFold2 models
The AlphaFold2 models of individual domains were extremely similar to the cryo-EM-derived structures. The only notable disagreement of AlphaFold2 models with experimental data was in several regions of subunit contacts some of which are shown in Fig. 4. The superposition of cryo-EM-derived structures and AlphaFold2 models resulted in the following root mean square deviations between the equivalent Cα atoms: 3.08 Å (465 out of 484 atoms) for gp105, 2.00 Å (628 out of 649 atoms) for gp089, 2.50 Å (417 out of 426 atoms) for gp270, 2.42 Å (629 out of 631 atoms) for gp154, 1.54 Å (256 out of 261 atoms) for the N-terminal domain of gp226 (gp226 NTD), and 2.76 Å (169 out of 169 atoms) for the C-terminal domain of gp226 (gp226 CTD). Notably, the AlphaFold2 models were excellent at predicting the structure of dynamic peripheral domains (e.g. the NTD and CTD of gp226). Additionally, the AlphaFold2 helped to identify several cis prolines, which significantly improved the geometry of the surrounding regions.