Figure 4.

Reconstruction Model Quality and Consistency

(A) The clustering analysis of 10 reconstructed models for dataset SASDA25. Using the correlation distance cutoff of 0.3, 10 reconstructed models were classified to two groups (see [C] for the reconstructed models superposed to the atomic structure of the same protein).

(B) The theoretical SAXS profiles of reconstructed models were fitted to the experimental data.

(C) The 10 reconstructed models grouped into two classes, enclosed by the red and green boxes, corresponding to the two classes in (A). The correlation coefficients compared with the PDB structure are indicated next to each reconstructed model.