qindex

qindex is an experimental program to do initial screening of 2-D crystals, and perform single-particle style averaging on data from poorly ordered crystals. It was developed in collaboration with Gina Sosinsky (as an end-user). This is not exhaustive documentation, but should be enough to get you started with it:

• run qindex, open a micrograph (real-space image)
  • note, please convert your micrograph to mrc format with 'proc2d' before opening it. It is necesary to be in this format in one step.
• 4 windows will appear
  • control panel
  • window containing the full micrograph you just opened (pan using RMB or panning widget in the control panel, middle click as usual for more controls) Drag with the LMB to select an area of interest.
  • window containing a boxed out section of the micrograph. The size of the region is specified in the control panel as Size. A grid corresponding to the current unit-cell spacings is overlaid on this window.
  • window containing the FFT of the boxed out region with an overlaid unit cell in Fourier space. Enter the A/Pix value in the control panel, then drag any of the light-blue circles in this window to define the unit-cell (if you use a higher order circle, you will get more precise positioning. Unit cell parameters are shown in the control panel. Note that you can use the middle button in this window as well…
  • You can read out specific peak intensities from the Peak subwindow in the control panel.

There are then 5 buttons next to the panning widget. These are basically designed to be used sequentially (except clear)

1. Pick a good area of your crystal and adjust the unit cell parameters 2. press 'Avg' (this may take some time depending on the unit cell size) 3. a new window will appear with an improved contrast average about 2-3 unit cells in size (varies a bit). Hoepfully the real-space crystal will appear very visibly in this window, but how well it works will depend on the contrast a bit… 4. set a 'ref size' in pixels corresponding to the size of the region you want to 'reconstruct' (in 2d), generally a couple of unit cells, then left-drag in the new window to select your phase origin 5. click on the 'corr' button. This will cross-correlate the average region you just selected with the entire micrograph, giving you an overview of crystal quality across the image. Of course this will only be accurate within a single crystal. This replaces the original micrograph. 6. raise the display contrast in the micrograph window to 1.0, then adjust the brightness so you can see a pattern of spots corresponding to good regions of your crystal. 7. press the 'peak' button, this will apply a peak filter to the correlation map. Readjust the brightess. Any peak you see on the screen will be treated as a 'good' region for further processing. 8. press the 'box' button. The number of selected dots will appear below the panning widget, and a box database fill will be written to disk with the name 'qindex.box'. 9. quit qindex (or hide it)

batchboxer input=<your micrograph> dbbox=qindex.box output=boxed.hed
  (you can use the newsize= option if you want to change the box size)
proc2d boxed.hed boxed.avg.hed average
v2 boxed.avg.hed

you can experiment with other ways of averaging as well, including alignment

Contact me (Steve Ludtke) if you need more assistance.