This is a small collection of scripts which help with cryoCARE denoising of cryo-ET data. In short, frames will be sorted into even and odd, aligned using MotionCor2, manually reconstructed using imod, and finally denoised using cryocare.
- JugLab for cryoCARE
- Zheng et al for MotionCor2
- Mastronarde and Held for IMOD
-
Run motioncor2.sh script to align frames and create EVN & ODD tilt series
- Install MotionCor2 as described here: https://msg.ucsf.edu/software
- Update the motioncor2 installation path in the script
- Frames need to be saved as .tif files with the following naming scheme: TS_aa_bbb_cc.c.tif
- with aa=tilt series, bbb=serialEM object and cc.c=tilt angle.
- Raw data needs to be organized in a specific way in order for the script to be functional:
- ./01_raw_data <-- in this folder all realigned and ordered .st/mrc and .st/mrc.mdoc files are stored
- ./01_raw_data/frames <-- in this folder all .tif files are stored
- ./motioncor2.sh <-- this is the correct location for this script
- This script will generate a hidden temporary folder called .motioncor2_temp where all intermediate files are stored.
- The final aligend files will be saved in the folder: ./02_motioncor2
-
Reconstruct tomogram
- Manually reconstruct tomogram using etomo
-
Reconstruct even / odd tomograms
- Copy the folder of the original reconstruction in two new folders
- Replace the tilt series with the even / odd files
- Reconstruct again with identical parameters.
-
Run cryoCARE
- Install cryoCARE as described here in a conda environment: https://github.com/juglab/cryoCARE_pip
- Create a new folder for cryoCARE
- Copy all necessary files into the folder:
- predict_config.json
- train_config.json
- train_data_config.json
- cryocare.sh
- The two reconstructed tomograms, they should be named EVN.mrc and ODD.mrc
- Activate the conda environment: conda activate cryocare
- Run the cryocare script: ./cryocare.sh