Please also check the RapiData2012 SHARP/autoSHARP Wiki!

Please check here from time to time - we'll update it during the course as we go along! Come and ask us (Claus & Clemens) if you have any questions.



The autoPROC software should be installed on all machines used during the workshop. At the moment (24th April 2012) you still need to type

% autoproc

(to setup the environment) before you can use the commands. Afterwards, you can test that everything is setup correctly with

% process -h
% aP_scale -h

The main commands for you to use (which all have help available when run with the –h flag):


Using your own data

Here are some suggestions for getting started on your own data very quickly - this is not a substitute to going through the full set of tutorials though.

The most simple form is to run

% process -d 01 | tee 01.lis

in a directory with images. To keep output organised we use the –d flag (to have all output go into a subdirectory) and save the standard output (with important notes and warnings) as well.

Alternatively, you could run in any directory and specify

% process -I /where/ever/images/directory -d 02 | tee 02.lis

If you think your data is good you could also try

% process -M automatic -d 03 | tee 03.lis

What are the things to look out for? Have a look at any warning messages in standard output - they often point to potential problems.


If indexing fails it is possible that the beam centre is either wrong or using a different convention from what is expected by autoPROC, try running with

% process BeamCentreFrom=getbeam:init -d 04 | tee 04.lis

Or maybe the rotation axis has the opposite rotation sense? then try

% process ReversePhi=yes -d 05 | tee 05.lis

You could prime autoPROC with a known cell and/or spacegroup:

% process cell="20 30 40 90 110 90" symm=P2 -d 06 | tee 06.lis

Or if the pattern is rather messy (suggesting multiple lattices) using the 'iterative indexing' feature with

% process XdsOptimizeIdxrefAlways=yes -d 07 | tee 07.lis


Sometimes, integration can have unstable parameter refinement. This can be seen by checking the files distance.png, detector_center_origin.png, cell_axes_devmean.png and standard_deviation.png. It can be stabilised using

% process autoPROC_XdsKeyword_REFINEINTEGRATE="AXIS" -d 08 | tee 08.lis

If there is a serious amount of ice-rings visible (check SPOT.noHKL.png), exclude these already during integration with

% process XdsExcludeIceRingsAutomatically=yes -d 09 | tee 09.lis


It is easy to try different scaling approaches with the aP_scale module:

% aP_scale -mtz XDS_ASCII.mtz -P Lysozyme Soak1 peak -b 1-360 -id 01 | tee 01_aP_scale.lis

Or for a 2-wavelength MAD dataset something like

% combine_files -f peak/XDS_ASCII.mtz -P ProtKin SeMET peak \
                -f infl/XDS_ASCII.mtz -P ProtKin SeMET infl -o peak-infl.mtz
% aP_scale -mtz peak-infl.mtz -P ProtKin SeMET peak -b 1001-1720 \
                              -P ProtKin SeMET infl -b 2001-2720 \
           -id 02 | tee 02_aP_scale.lis

The important flags are –P (especially the last argument to it) and –b: these flags together with the order of them defines what sets of images are going to be merged (they will be scaled together anyway).

Data provided for 'playing' by organisers

The organisers of the workshop will provide some data for 'playing' - especially useful during the start of the workshop when no new data has yet been collected.

Available are

Sets of images are in


and can easily processed - e.g. with

% mkdir /img06/data1/pxuser/RD2012/play/MyName/test1
% cd /img06/data1/pxuser/RD2012/play/MyName/test1
% process -I /img03/data1/pxuser/RD2012/ly49bp-4/6 -d 01 | tee 01.lis

See the SHARP/autoSHARP wiki page for details about solving those structures.

Data from previous RapiData courses

Using data provided during previous RapiData courses:

Other tutorials