view article

Figure 5
The cryoDRGN (Zhong et al., 2021BB48) Flexibility Hub workflow to analyse the continuous flexibility of the SARS-CoV-2 spike in the 1Up and 3Down conformations. (a) shows the current integration of the cryoDRGN pipeline in the Flexibility Hub, which includes two main steps: particle preparation and neural network training. The resulting conformational landscapes captured by the latent space of the trained network can subsequently be inspected and annotated with the Flexutils interactive tools. (b) provides an example of the 2D interactive annotation tool integrated into Flexutils. The viewer shows a hexbin plot resulting from a PCA reduction of the original latent space computed by cryoDRGN. Two main groups can be identified in this view corresponding to the 1Up and 3Down conformations of the spike. In addition, five points were selected along the main variability axis of the 3Down point cloud region. (c) shows an example of a cryoSPARC (Punjani et al., 2017BB33) homogeneous refinement with the closest particles associated with one of the five picked points (middle point) illustrated in (b). For this test, the 5000 original particles closest to the picked point were registered in Scipion, which led to a reconstructed volume with a resolution of 4.1 Å.

Journal logoSTRUCTURAL
BIOLOGY
ISSN: 2059-7983
Follow Acta Cryst. D
Sign up for e-alerts
Follow Acta Cryst. on Twitter
Follow us on facebook
Sign up for RSS feeds