relion-classranker

The class ranker is part of the cryogenic electron microscopy (cryo-EM) dataset processing pipeline in RELION. It is used to automatically select suitable particles (EM images) assigned to 2D class averages for further downstream processing. This model comprises two main components: a CNN responsible for extracting image features, and an MLP that combines the image features with additional statistics to assigns a score ranging from zero to one for each 2D class average. The selection is subsequently done in RELION through a user defined cutoff for the predicted score. For training, a supervised approach is adopted, where pairs of 2D classes and corresponding human-assigned scores are used to teach the model.

More Information

For more information visit the relion-classranker website.

Available Versions

These versions of relion-classranker are available on Baskerville. These will be retained in accordance with our Applications Support and Retention Policy.