In June 2023, we provided two single-particle datasets with continuous heterogeneity to the cryoEM methods development community. Our intention was to provide an opportunity to compare results of various heterogeneity analysis methods on identical datasets.
Here you will find the anonymised submissions to the heterogeneity challenge, ready for import into an analysis pipeline https://github.com/flatironinstitute/Cryo-EM-Heterogeneity-Challenge-1. We welcome feed back and contributions to this codebase.
The original may be downloaded at the following link
https://www.simonsfoundation.org/flatiron/center-for-computational-biology/structural-and-molecular-biophysics-collaboration/heterogeneity-in-cryo-electron-microscopy/
Dataset 1 is an experimental dataset of thyroglobulin, collected at the New York Structural Biology Center. Compositional heterogeneity was carefully screened out of this dataset, so particles only vary in the underlying conformation of the molecule. The poses and defocusses obtained by cryosparc refinement for this datset were used for the second dataset.
Dataset 2 is a synthetic dataset. The molecular structures for this dataset were sourced from course grained molecular dynamics simulations of thyroglobulin, and particle images were rendered with the use of the cisTEM simulator package. The structures of thyroglobulin were sampled to produce a non-trivial distribution of 3871 unique structures. Submissions will be assessed on how accurately they are able to able to recover this distribution.