X-Ray Data Processing

Prepare data

XDS can't work with h5-format images, so you need to convert them to cbf format before processing. On macOS, eiger2cbf is a useful tool for this.

eiger2cbf path/to/your/h5_master_data           -- get number of frames
eiger2cbf path/to/your/h5_master_data N out.cbf -- write N-th frame to out.cbf
eiger2cbf path/to/your/h5_master_data N         -- write N-th frame to STDOUT
eiger2cbf path/to/your/h5_master_data N:M   out -- write N to M-th frames to outNNNNNN.cbf

For example:

eiger2cbf path/to/your/h5_master_data 1:720 out # convert all the 720 images to cbf format

X-Ray data processing

Process with XDS

1. Install XDSGUI following these instructions.

2. Run XDSGUI in the terminal.

   BASHCopy

   cd path/to/your/images
   xdsgui
   

3. Load the image and generate the XDS.INP file:

   

4. Modify the XDS.INP file to meet your needs, then click Run XDS.

   :::caution

   If you collected your data at the 10U2 beamline at SSRF, change the ROTATION_AXIS parameter to 0 -1 0.

   :::

5. Check the results on the CORRECT page.

Process with Xia2

1. You can use Xia2 in two ways: within CCP4, or by installing the latest xia2/DIALS bundle as described on this website.

2. Use it to process your data:

   BASHCopy

   cd path/to/images
   xia2 pipeline=dials-aimless ./ goniometer.axes=0.000000,1.000000,0.000000 xia2.settings.resolution.d_min=2
   

Process with autoPROC

1. Request a license and install autoPROC as described on this website.

2. Use it to process your data:

   BASHCopy

   cd path/to/images
   process -I . -d autoproc -R 50 2 > out.log 
   #save the result in subdirectory named autoproc, resolution vary from 2 to 50, save the log to out.log file.
   

Structure determination

Validate the data

1. Use the XDS_ASCII.HKL file in the autoproc directory as the diffraction file.

2. Validate the data with Phenix Xtriage: load the diffraction file into it and run.

   

Molecular replacement

1. Use the XDS_ASCII.HKL file in the autoproc directory as the diffraction file.

2. Get the template model:

   1. BLAST your protein sequence against the PDB database, select a suitable structure, and download the PDB file.

   

   2. After downloading the template model, edit it in PyMOL to remove extra copies and water molecules.

3. Determine how many copies are in the ASU (asymmetric unit):

   1. You can get this number from the Xtriage output.

   2. Or use the cell content analysis tool in CCP4.

   Note

   You can determine the number of copies from the solvent content. Generally, the solvent content of most biological macromolecular crystals is between 40% and 60%.

4. Run molecular replacement with Phenix Phaser-MR.

5. Build your model with Phenix AutoBuild: load the molecular replacement result and run it.

   

6. Check the model from AutoBuild in Coot and refine it manually.