Chapter 2: Autoindexing and Cell Refinement
Step 12: 3D Window and Mosaicity
Step 12: 3D Window and Mosaicity
Set the 3D Window so that the Oscillation Range encompassed by this sector of reciprocal space is at least twice as large as the refined mosaicity. Typically a value of 5 frames is used for ccd detector data. For IP data you should keep the 3D Window at 1 initially, due to a high probability that consecutive frames are not scanned in the same way.
The 3D Window specifies the number of consecutive frames from which the reflections will be integrated simultaneously. Typically for CCD data a value of 5 frames is used. This is a new feature in Denzo. One principal advantage of this feature is that binning frames together can include more reflections in the refinement, thereby increasing the accuracy of indexing on frames that have very few reflections (e.g. small molecule data sets, low mosaicity / small unit cell macromolecular data sets, low resolution macromolecular data sets). The disadvantage of grouping frames together is that local variations in frames are averaged out (e.g. "bad" frames, frames where the crystal slipped, etc.)
For data collected on IP scanners (Raxis, Mar, DIP) errors in the reproducibly positioning the IP with respect to the scanning machinery preclude the use of a 3D Window greater than 1. That’s why the default value for the 3D Window is 1 for these sites.
Unlike CCD detectors, which are solid state, image plate detectors move the IP relative to the scanning head in the course of data acquisition, data read-out, and plate erasure. While these devices are very well made, there is, nonetheless, a small non-reproducibility in their operation that results in slight variations in the actual position of the direct beam and the Distance. Until it has been established that this variation is small relative to the pixel size, the 3D Window should be kept at 1 to avoid inaccurate mapping of the pixels of sequential frames onto each other.
The limitation for 3D Window usage is the primarily computer memory. All the frames of the 3D Window must be held in memory simultaneously. This is why it is recommended to have between 512 and 1024 Mbytes of memory for data processing computers.
For small molecule data collection, the reciprocal lattice is very sparsely populated and it is reasonable to set the 3D Window to 30 or more frames. This could easily consume 300 Mbytes or more of memory, depending on the frame size.
One principal advantage of the 3D Window is that binning frames together can include more reflections in the refinement, thereby increasing the accuracy of indexing on frames that have very few reflections (e.g. small molecule data sets, low mosaicity / small unit cell macromolecular data sets, low resolution macromolecular data sets). The disadvantage of grouping frames together is that local variations in frames are averaged out (e.g. "bad" frames, frames where the crystal slipped, etc.) A good rule of thumb is to set the 3D Window (after the resolution limits and lattice have been determined) to be at least twice the refined mosaicity. The default value of 5 is a reasonable compromise.
Mosaicity is defined in Denzo as the rocking angle, in degrees, in both the vertical and the horizontal directions, which would generate all the spots seen on a still diffraction photograph. It includes contributions due to X-ray bandwidth, beam crossfire, etc. Mosaicity is refineable when integrating your frames if the 3D Window is set to be at least twice as large as the refined mosaicity. For example, if your crystal has a mosaicity of about 1 degree, then a 3D Window encompassing 2 degrees or more of data should allow for proper refinement.
The histogram is the mosaicity histogram for the reflections of the frame or of the frames of the 3D Window. The reflections are sorted into 20 zones, which range from minus the input or refined mosaicity/2 to plus the input mosaicity/2 (Figure 39).
Each zone represents the shortest angular distance of the center of the reflection from the surface of the Ewald sphere at the end of the oscillation range, and is formally defined as average observed partiality of the reflections in each zone. In other words, if, at the end of the oscillation range a particular reflection was -0.3 degrees away from (i.e. past) the surface of the Ewald sphere, then it would contribute to the “bar” in the -0.3 degree zone.
Only reflections that are single partials are included in the analysis. A reflection in negative zones means that the center of the reflection has already passed through the Bragg condition. In all cases, no matter what mosaicity you have input to Denzo, the histogram should pass through 50% in the zero zone. If the mosaicity you guessed/input was too small, then the histogram will not descend to zero at +mos/2. If the mosaicity you guessed/input was too large, then you will see a histogram which looks like a step function, falling to zero just past the zero zone.
If the mosaicity was chosen correctly or slightly overestimated (which is the preferred side to err on) the histogram will descend smoothly to zero at +mos/2. Frequently this is not strictly the case and the histogram tails off due to the effects of diffuse scattering. If there are only a few reflections, the histogram may be quite choppy. The number of reflections can be increased by increasing the size of the 3D Window. If the histogram does not resemble any of the shapes mentioned here, it may be indicative of motor, spindle, or shutter problems.
Figure 39. The mosaicity histogram
The best way to set the mosaicity is to use the 3D Window and let Denzo refine it in the course of integrating the reflections. If you are unable to use a 3D Window greater than 1, and the mosaicity is greater than half your oscillation range (which is usually the case, e.g. 1 degree oscillations, 1 degree mosaicity), then you will have to set the mosaicity manually. You do this by going to the Crystal Information button and entering the Mosaicity in the field provided.
A correctly set mosaicity will not always result in preds covering all of the reflections on an image. In particular, very strong low order reflections may require unreasonably large values of the mosaicity for coverage. These reflections may not represent the intrinsic mosaic spread of the crystal, but, rather, may result from particularly intense diffraction from that lattice point.
A mosaicity that is set too high is better than one that is set too low, but setting the mosaicity too high could result in an unnecessarily high number of overlaps.
You do this by going to the Crystal Information tab (Figure 40) and entering the Mosaicity in the field provided.
Figure 40. The Crystal Information window
New lattice refinement
Table of Contents