[3dem] Automatic 2D class selection with Cinderella

[Thorsten Wagner]

Dear colleagues!

  "The good ones go into the pot, the bad ones go into your crop."

We are happy to announce the release of Cinderella, an automatic 2D class selection tool: http://sphire.mpg.de/wiki/doku.php?id=auto_2d_class_selection

Cinderella is based on the same deep convolutional neural network as crYOLO and was trained on classes of 17 datasets (number increasing!) to separate good from contamination classes. The program is open source and standalone.

Cinderella is easy to use. You can use our pretrained network or easily train your own class selection network to separate classes more specifically (i.e. based on heterogeneity) . It supports .hdf and .mrcs files for training / classification.

We uploaded all published training data and invite the community to contribute additional classes to improve the pretrained network even further:
http://sphire.mpg.de/wiki/doku.php?id=auto_2d_class_selection#contribute

With Cinderella, SPHIRE is providing the next tool for hands-free processing of cryo-em data.

We hope you will like it. Try it out!

If you have further questions, please contact us at: https://listserv.gwdg.de/mailman/listinfo/sphire

On behalf of the SPHIRE development team
Thorsten, Luca, and Stefan


_____________________________
Dr. Thorsten Wagner
Max-Planck-Institute of Molecular Physiology
Structural Biochemistry
Otto-Hahn Strasse 11
D-44227 Dortmund
Phone +49-(0)231-133-2357
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.ncmir.ucsd.edu/pipermail/3dem/attachments/20190529/9c9c1664/attachment.html>