Basic Usage of xia2.multiplex

Tutorial Data

The data (~6GB total) were collected on I24 at Diamond Light Source and is available to download from cows_pigs_people_data. 36 insulin crystals were collected from a range of 🐮, 🐷, and 💁‍♀️, with small rotation wedges recorded on each. All data have symmetry I 21 3. For this introduction to multiplex, we will only be using the data from the 🐮 insulin (~2GB). For a more advanced tutorial in automatically handling multi-crystal data with multiple “things” present, see this tutorial.

Once you have the data downloaded, this script (linux / UNIX) can help extract the data faster.

mkdir data
cd data
for set in CIX1_1 CIX2_1 CIX3_1 CIX5_1 CIX6_1 CIX8_1 CIX9_1 CIX10_1 CIX11_1 CIX12_1 CIX14_1 CIX15_1 ; do
wget https://zenodo.org/records/13890874/files/${set}.tar
tar xvf ${set}.tar
rm -v ${set}.tar
done

DIALS Workflow

xia2.multiplex is a tool to scale and merge multi-crystal data: therefore, it requires integrated files to start with. To get our integrated files from our raw images, we can use a standard DIALS workflow. Note that data from multiple crystals will not in general share an orientation matrix, so the indexing should not join all the lattices.

This tutorial assumes knowledge of processing with DIALS. For more information, see this DIALS tutorial.

dials.import data/*gz
dials.find_spots imported.expt
dials.index imported.expt strong.refl joint=False
dials.refine indexed.expt indexed.refl
dials.integrate refined.expt refined.refl

xia2.multiplex

xia2.multiplex can now be run on this data. First, the software will do some preliminary unit cell clustering to remove any datasets with a very different unit cell. Next, consistent symmetry will be applied and any indexing ambiguities are resolved. Note that this uses dials.cosym under the hood. Next, the data is scaled, and the space group is determined. Once the data is merged, a range of analyses are performed which are useful for diagnostics. Let’s explore this on the example data.

xia2.multiplex integrated.expt integrated.refl

Note that if you have separate integrated files you can list them in the input command.

xia2.multiplex integrated_1.expt integrated_1.refl integrated_2.expt integrated_2.refl
There are also some additional command-line options you may find useful:

  • symmetry.space_group - set the space group if known, which can speed up symmetry analysis.

  • resolution.d_min - set the resolution cut-off, otherwise it will be automatically determined.

  • filtering.method - applies additional filtering, see here.

  • clustering.output_clusters - output identified sub-clusters, see here for more choices and details.

Diagnostic Tools in xia2.multiplex

Most diagnostic graphs are available in xia2.multiplex.html. Some extra plots are output as .png files to the working folder.

Unit cell analysis: The unit cell of each crystal input to xia2.multiplex is plotted to give an initial analysis of isomorphism. To see the unit cell distributions of all input data, open cluster_unit_cell_p1. If you see subgroups in the dendrogram denoted by different colours, this means there are some major differences in unit cell and only the largest will be considered for analysis with xia2.multiplex. If the entire dendrogram is the same colour, then all datasets will be fed into the pipeline. The xia2.multiplex.html log file also contains graphs related to unit cell clustering. It is important to note that these graphs only pertain to any subset chosen by the initial unit cell filtering. Subtle differences in unit cell can still occur, and the initial unit cell filtering is fairly conservative, so these graphs are worth examining. In this case, all the data were accepted during the initial filtering, and are shown to have very similar unit cells.

../_images/multiplex_uc.png

Orientation analysis: Multi-crystal data collections often have small wedges either due to radiation damage constraints, or sample environment constraints (i.e. in situ data collections). A risk with such experiments are that all the crystals are measured in the same orientation, and the same small volumes of reciprocal space are measured repeatedly. Coverage of reciprocal space can be assessed by considering the distributions of the unit cell axes in reciprocal space. A stereographic projection plot is used for this. If the dots are widely distributed around the circles, then there is wide coverage of reciprocal space and preferential orientation is not an issue.

../_images/multiplex_sp.png

\({\Delta}CC1/2\): Another useful tool is the \({\Delta}CC1/2\) analysis. This shows whether individual datasets contribute positively, or negatively to the overall quality of the merged dataset. Note that any outliers are not removed unless filtering.method=deltacchalf is used. In this case, the \({\Delta}CC1/2\) for each dataset is quite minimal, although one dataset has a stronger effect than the others and could be removed if desired. The intensity clustering can also be compared to help determine if outliers here are significant.

../_images/multiplex_dcc.png

Intensity clustering: An analysis that compares the intensities of each dataset to see if there are any differences. This is useful to identify if there are subsets within the data that should be grouped together and merged separately. See this tutorial for more information on interpreting the graphs within this section.