Fig. 3: Consensus landscapes obtained by FlexConsensus for the EMPIAR 10028 dataset14.
From: Merging conformational landscapes in a single consensus space with FlexConsensus algorithm
Three conformational spaces were input to the FlexConsensus analysis, obtained from three independent runs of the following methods: HetSIREN in reconstruction/refinement mode5 and Zernike3D6. Panel a compares the common consensus landscape (in grays) against each consensus subspace generated from each input (colored). The comparisons show a region identified by both HetSIREN runs that is not present in Zernike3D, corresponding to compositional changes of the ribosome. It should be noted that, due to the approximation of each method, HetSIREN can identify compositional variability, whereas Zernike3D focuses on continuous conformational changes. Panel b displays the three subspaces presented in a individually, without being superimposed on the common consensus space. Panel c highlights the regions assigned a higher consensus error for each method. For Zernike3D, the highlighted regions correspond to the region where the biggest compositional variability change has been detected by HetSIREN, which is expected, as Zernike3D cannot properly identify compositional variability. In the case of HetSIREN, these regions are located on the border regions of the 3D consensus landscape that are farther from the main cloud. rec., reconstruction; ref. refinement.
