Fig. 5: Performance of atomic model fitting.

a Comparison of RMSD of the alignment with the best score by VESPER (DOT) with CC, gmfit, and fitmap. EM maps in the dataset are listed in Table 3. A voxel spacing of 3 Å and a rotation angle of 10° were used in Fig. 4. Supplementary Fig. 4 provides results for the other three parameter combinations. Blue circles, comparison against CC; orange triangles, gmfit; green crosses, fitmap, respectively. b The fraction of query chains for each map that had the top-scoring alignment with an RMSD of 5.0 Å or less (solid gray bars) and 10.0 Å or less (including hatched bars). Black bars, VESPER (DOT); dark gray, CC; medium gray, gmfit; pale gray, fitmap. c Chain structure fitting for five chains of γ-secretase (EMD-3238, PDB ID: 5fn3). The left panel shows superimposed structures of 5fn3 in the map. Each chain is in different color. From the second to the last panel from the left, chains placed within an RMSD of 5.0 Å are shown for VESPER (DOT), CC, gmfit, and fitmap, respectively. The number of chains placed within the cutoff was 4, 1, 1, and 2 chains by these methods, respectively. d Chain structure fitting of transcription factor IIH (EMD-3802, PDB ID: 5of4). There are 10 chains to fit. VESPER (DOT), CC, gmfit, and fitmap placed eight (A, B, E, F, H, X, Y, Z), three (A, B, Y), one (B), and zero chains within 5.0 Å RMSD, respectively. Chain IDs are taken from the PDB file. For fitmap, the placement of chain B, which had an RMSD of 37.7 Å is shown, since this chain had the smallest RMSD among the other chains. e fitting of 18 chains of RNA polymerase I-Rrn3-CF complex (EMD-3591, PDB ID: 5n5z). Within 5.0 Å RMSD, VESPER (DOT), CC, gmfit, and fitmap placed seven (A, B, C, E, G, H, O), four (A, B, C, O), zero, and two (B, P) chains, respectively. For gmfit, chain A that was fit at an RMSD of 9.0 Å is shown.