This walkthrough of cryoDRGN analysis of the assembling ribosome dataset (EMPIAR-10076) from Figure 5 of Zhong et al. covers:

1. preprocessing of inputs,
2. initial cryoDRGN training and analysis,
3. removing junk particles,
4. high-resolution cryoDRGN training and analysis,
5. extracting particles for traditional refinement, and
6. generation of trajectories.

Figure 5 from Zhong et al 2021.

SI Video 3 from Zhong et al 2021

SI Video 3 from Zhong et al 2021

• For an abbreviated overview of the steps for running cryoDRGN, see the README on the github page: https://github.com/zhonge/cryodrgn
• All pre-processed inputs that are generated in this tutorial (except particle images) are also located here: https://github.com/zhonge/cryodrgn_empiar
• The final trained networks and density maps are deposited here: https://zenodo.org/record/4355284.
• For compute requirements and installation instructions, see cryoDRGN installation with anaconda.

Preparing cryoDRGN inputs

Step 1) Obtain the dataset

• Download EMPIAR-10076 (~51GB) from https://www.ebi.ac.uk/pdbe/emdb/empiar/entry/10076/

  • You can download it directly from the browser, or from the command line using Aspera Connect (More info here)
    • $ ascp -QT -l 200M -P33001 -i ~/.aspera/connect/etc/asperaweb_id_dsa.openssh [[email protected]](<mailto:[email protected]>):/10076 .

• This dataset contains 131,899 extracted particles with box size of 320 and pixel size of 1.31 A/pix.

  

  Screenshot from the EMPIAR entry

• The particle images are in the file L17Combine_weight_local.mrc. There is a typo in the file extension, so rename this file with the .mrcs extension.

  • $ mv L17Combine_weight_local.mrc L17Combine_weight_local.mrcs

• The CTF parameters for each particle are in the metadata file, Parameters.star. We will extract the CTF parameters from this file later in Step 3.

  • $ head Parameters.star -n 20

• Poses (i.e. particle alignments) are not present in the deposited data, so we will next run a consensus reconstruction.