Structure Determination and Modelling

Crystallography Software

Data collection procedure

Fig. 2.1 - Click to enlarge

Double click CrystalClear 1.3.5 icon on the desktop. The login screen with CrystalClear logo is appeared.Crystal Clear program opens with a full screen (see Fig. 2.1). This screen has several sections: 1) toolbar section, 2) task window 3) icon section, 4) flow bar window, 5) messages window, 6) Stop/Abort button, and 7) Info window.

Following the full screen window, crystal clear will prompt for a new project and/or new sample name, depending upon whether it is a new project or a new sample.

Initialize Instrument begins immediately. If you click yes, the instrument will be initialized.

In Setup Menu, ensure the following five fields are correct

Crystal to Detector Distance

Detector 2q

X and Y Direct  Beam Positions

Source wavelength

In the Task dialog, there is the possible selection includes: Screen Collect and Process, Collect and Process, Process, etc. Depending upon the nature of data collection select the appropriate task, for example, if you want only collect the data, select Collect. But if you wish to first screen several samples before collection, select Screen Collect and Process.

In The Image Files dialog, you could see your default data directory. If this is incorrect, use the Browse button and select the correct directory for the images. Finally click Finish.

Crystal Evaluation is pre-collection-phase operation. Before collecting Initial Images, you will mount crystal. Remember locking the phi-axis. Initial Images is used to collect a small sample of images to evaluate the crystal. Up to 20 images can be collected.
In the Collect Initial Images dialog, you may adjust the Scan Table. Click Run to start the scheduled collection.

Assign Unit Cell is used to determine and confirm the unit cell and mis-setting angles.

Find Spots is the first step in Assign Unit Cell operation.. It is recommended using 50-70 reflection centroids  The exact number needed can vary depending on the centroid accuracy and location in reciprocal space. Adjust parameters as desired.

Index spots indexes reflection centroids in the selected reflection list to elucidate the crystal unit cell dimensions, the crystal orientation, and the Bravais Lattice type. If desired, change parameter values in the Advanced tab.

Select a solution by double clicking on the row in the Index Results Window and clicking the OK button.

Refine Cell allows the user to refine the crystal, detector, and source parameters in order to reduce the differences between observed and calculated reflection centroids. The refinement process is complete when the number of cycles are completed. If desired, click the advanced button to change the parameters, and click run to begin.

Predict spots is typically used to confirm the results of the Refine Spots operation. It provides the user a mechanism to determine a good estimate for crystal Mosaicity, as would occur if 2D data were used for indexing.

Strategy can be used to automatically generate a collection schedule so the user doesn't have to collect manually.

Collect Images is useful in working with Schedules. New collection schedules or screen schedules can be created, run and saved. To watch the process of the collection, click the Show Instrument State Display button. Click Run to use a Schedule to collect images. Review the specified settings on the detector. Click the Start Image Collection.

The Integrate Reflections operation predicts the reflections that would appear on a range of images. Integrate Reflections can be executed once the images have been accurately predicted. This operation may produce several reflection files that are merged as part of the integration.

Once integration is completed, you may wish to analyze the data for purposes of evaluating and determined Laue symmetry, Centricity, and Space group.

Scale calculates and applies scale factors to different batches of reflections in a reflection list, averages symmetry equivalent reflections, calculate merging and completeness statistics, and creates a reflection list of unique reflections. The Scale and Average Results window displays graphs which depict the scaling results, generates a log file, and writes the reflection file to the sample directory.

XdisplayF is used to display the images. Denzo_3D is used to auto-index, refine and integrate the images, and Scalepack is used to scale all of the observations to get one data set.

1, run HKL2000 with GUI

Unix prompt: HKL2000 (note capitals)

Set up your data files in the data menu. Identify where the raw images are and where to save the results.

Note: the beam position, crystal to detector distance, and oscillation range for image (and data)

First, pick spots on the image to be used for indexing.

Move to the indexing tab/window. Display image and search for peaks (rule of thumb ~300 spots are good for indexing). Use the peak pick to add or subtract peaks in the image.

Click index in the indexing window. The file peaks.file has the x and y coordinates of each peak that was identified. Poorly diffracting crystals may not have lots of peaks in one image; you can add more than one image using the frame command in the image display window in order to use the peaks from multiple frames for indexing.

Indexing: determine the space group (actually the point group). The lower the score the better the fit. Choose the cell with the most symmetry and a low score. Note that the best score is always P1. Click on the best cell and click on save and close.

Based on the chosen cell, we will predict where the spots should be. We will evaluate whether we agree with our first guess. These parameters will be used to predict where all the spots should be on each successive frame in the data set.

First use all the data in this process. If it doesn¡¯t work use the step resolution method. Set the resolution to ~3.0 angstroms and FIT BASIC, click REFINE.

The yellow (or green) rings should be around the spots (yellow = partial, green = fully measured).

First, make sure you are getting all the spot in your integrating circle (can also use an ellipse). Zoom window, int box, etc. The goal is get all the spot and get a good area for background corrections. Change the size of the spot, background and boxes so that the box is covering the spot correctly.

REFINE until all parameters have convered and chi*2 have bottomed out.
Increase resolution in steps until max for your data and REFINE at each step.

FIT ALL (except mosaicity) and  REFINE in order to minimize the chi*2 values (difference in predicted and observed spots, below 2).

FIT mosaicity (and all other values) and REFINE.


Go to integrate page. Change file names, choose resolution, etc Scale sets.

Check Chi squared for all the data. It should be close to one. If it is under 1 then we are overestimating our errors and error scale factor should be altered. If over 1 then we have underestimated our errors. This is a reciprocal value, increasing ERROR SCALE FACTOR will reduce your chi*2 values.


Run Scaling twice (at least). The first is to write the reject file, the second is to remove the rejected reflections.

2, run HKL2000 without GUI

Denzo and Scalepack can be called in the standard way by typing denzo_3D or scalepack at the command prompt. Xdisp can also be called by typing xdisp at the prompt followed by the image name.

Denzo. in sample file example file example file

How to run and convert the reflection file

1. Convert the reflection File

This is simply to illustrate the procedure for getting scalepack data into CCP4, converted to amplitudes, add free flag, then CNS.

A script from Dr Maria Bewley

2. Run CNS

Before running CNS, you can edit your input files via the web browser interface.

To run CNS, type simply cns_solve at the command prompt followed by the input file name and an output file. i.e.:

%cns_solve < generate.inp > generate.out

a). generate.mtz (pdb), minimize.pdb, bindividual.pdb

The output files of generate.inp, minimize.inp and bindividual.inp. They are used as the input files for iterative minimization/b-factor refinement.


The ORIGINAL CV format data file, and the locally scaled file used by most refinement scripts

c). start.pdb, start.ion.pdb, start.ligand.pdb, start.water.pdb

The input files for generate containing your protein model, the ions, any ligand, and the waters.

d), ligand.param

These ligand parameter and topology files must exist in your working directory. They may be empty if you do not have a ligand.

  • A script that makes a ribbon diagram
  • A script that makes a ball-and-stick diagram
  • A script that incorporates an electron density map