❄️ 🐉 cryoDRGN reconstruction of EMPIAR datasets

This repository contains input files and commands for reproducing the cryoDRGN heterogeneous reconstruction experiments in Zhong et al.

Each directory describes one of the analyzed datasets and includes all required inputs for training excluding the particle images, which may be downloaded from EMPIAR. The commands provided here are compatible with cryodrgn version 0.3 and have been slightly modified to use updated best practices (e.g. --zdim 8 instead of --zdim 10, real-space particle windowing) and lead to qualitatively identical results (to our knowledge). The exact settings used in the original study can be found in config.pkl file associated with the trained models.

The outputs of the experiments (trained models and reconstructed volumes), which are too large to put in a github repository, can be found on zenodo.

EMPIAR-10028: Pf80S ribosome

Primary reference: Wong, W. et al. Cryo-EM structure of the Plasmodium falciparum 80S ribosome bound to the anti-protozoan drug emetine. Elife 3, e01963 (2014).

Pre-procesing inputs

$ cat empiar10028/inputs/README
# This directory contains all input files for cryodrgn training except for the particle stack from EMPIAR-10028

# Download EMPIAR-10028 particles (~51GB)
ascp -QT -l 200M -P33001 -i ~/.aspera/connect/etc/asperaweb_id_dsa.openssh emp_ext3@hx-fasp-1.ebi.ac.uk:/10028 .

# Downsample dataset (~26GB)
cryodrgn downsample shiny_2sets.star --datadir 10028/data -D 256 -o particles.256.mrcs -D 50000

# Extract pose and ctf information from cryoSPARC refinement
cryodrgn parse_ctf_csparc cryosparc_P11_J4_003_particles.cs -o ctf.pkl
cryodrgn parse_pose_csparc cryosparc_P11_J4_003_particles.cs -D 360 -o poses.pkl

Running cryoDRGN

$ cat empiar10028/run.sh
PARTICLES=inputs/particles.256.txt
POSES=inputs/poses.pkl
CTF=inputs/ctf.pkl
IND=filtered.ind.pkl
OUTDIR='outputs' # rename as desired

# To run with cryodrgn version 0.3.0
cryodrgn train_vae $PARTICLES --poses $POSES --ctf $CTF --ind $IND -o $OUTDIR --zdim 8 --enc-dim 1024 --dec-dim 1024 --amp -n 25 > run.log &

EMPIAR-10049: RAG1-RAG2 complex

Primary reference: Ru, H. et al. Molecular Mechanism of V(D)J Recombination from Synaptic RAG1-RAG2 Complex Structures. Cell 163, 1138–1152 (2015).

Pre-procesing inputs

$ cat empiar10049/inputs/README
# This directory contains all input files for cryodrgn training except for the particle stack from EMPIAR-10049

# Download EMPIAR-10049 particles (~66 GB), then move the downloaded .mrcs files to this directory
ascp -QT -l 200M -P33001 -i ~/.aspera/connect/etc/asperaweb_id_dsa.openssh emp_ext3@hx-fasp-1.ebi.ac.uk:/10049 .
mv 10049/data/ragSEC_150309.mrcs .
mv 10049/data/ragSEC_150311.mrcs .

# Extract pose and ctf information from cryoSPARC refinement
cryodrgn parse_ctf_csparc cryosparc_P53_J26_006_particles.cs -o ctf.pkl
cryodrgn parse_pose_csparc cryosparc_P53_J26_006_particles.cs -D 192 -o poses.pkl

Running cryoDRGN

$ cat empiar10049/run.sh
PARTICLES=inputs/allimg.star
POSES=inputs/poses.pkl
CTF=inputs/ctf.pkl
OUTDIR='outputs' # rename as desired

# To run with cryodrgn version 0.3.0
cryodrgn train_vae $PARTICLES --poses $POSES --ctf $CTF -o $OUTDIR --zdim 8 --enc-dim 1024 --dec-dim 1024 --amp -n 25 > run.log &

EMPIAR-10076: Assembling bacterial 50S ribosome

Primary reference: Davis, J. H. et al. Modular Assembly of the Bacterial Large Ribosomal Subunit. Cell 167, 1610--1622.e15 (2016).

Pre-procesing inputs

$ cat empiar10076/inputs/README
# This directory contains all input files for cryodrgn training except for the particle stack from EMPIAR-10076

# Download EMPIAR-10076 particles (~51GB), then move the downloaded .mrc file to this directory
ascp -QT -l 200M -P33001 -i ~/.aspera/connect/etc/asperaweb_id_dsa.openssh emp_ext3@hx-fasp-1.ebi.ac.uk:/10076 .
mv 10076/data/L17Combine_weight_local.mrc L17Combine_weight_local.mrcs

# Downsample images to D=256 (~33GB)
cryodrgn downsample Parameters.star -D 256 -o particles.256.mrcs --chunk 50000

# Extract pose and ctf information from cryoSPARC refinement
cryodrgn parse_ctf_csparc cryosparc_P4_J33_004_particles.cs -o ctf.pkl
cryodrgn parse_pose_csparc cryosparc_P4_J33_004_particles.cs -D 320 -o poses.pkl

Running cryoDRGN

$ cat empiar10076/run.sh
PARTICLES=inputs/particles.256.txt
CTF=inputs/ctf.pkl
POSES=inputs/poses.pkl
IND=inputs/filtered.ind.pkl
OUTDIR='outputs' # rename as desired

# To run with cryodrgn version 0.3.0
cryodrgn train_vae $PARTICLES --poses $POSES --ctf $CTF --ind $IND -o $OUTDIR --zdim 8 --enc-dim 1024 --dec-dim 1024 --amp -n 50 > run.log &

EMPIAR-10180: Pre-catalytic spliceosome

Primary reference: Plaschka, C., Lin, P.-C. & Nagai, K. Structure of a pre-catalytic spliceosome. Nature 546, 617–621 (2017).

Pre-procesing inputs

$ cat empiar10180/inputs/README
# This directory contains all input files for cryodrgn training except for the particle stack from EMPIAR-10180

# Download EMPIAR-10180 particles (~127GB), then move the downloaded .star file to this directory 
ascp -QT -l 200M -P33001 -i ~/.aspera/connect/etc/asperaweb_id_dsa.openssh emp_ext3@hx-fasp-1.ebi.ac.uk:/10180 .
mv 10180/data/Example/consensus_data.star .

# Downsample to D=256 (~80GB)
cryodrgn dowsample consensus_data.star --datadir 10180/data -D 256 -o particles.256.mrcs --chunk 50000

# Extract pose and ctf information from RELION consensus refinement
cryodrgn parse_pose_star consensus_data.star -D 320 -o poses.pkl 
cryodrgn parse_ctf_star consensus_data.star -D 320 --Apix 1.7 -o ctf.pkl 

Running cryoDRGN

$ cat empiar10180/run.sh
PARTICLES=inputs/particles.256.txt
CTF=inputs/ctf.pkl
POSES=inputs/poses.pkl
IND=inputs/filtered.ind.pkl
OUTDIR='outputs' # rename as desired

# To run with cryodrgn version 0.3.0
cryodrgn train_vae $PARTICLES --poses $POSES --ctf $CTF --ind $IND -o $OUTDIR --zdim 8 --enc-dim 1024 --dec-dim 1024 --amp -n 50 > run.log &