From: Harry Powell
Date: 16 April 2012 14:28
Dear all
We are pleased to announce the public release of new versions of iMosflm and Mosflm. We have addressed many bugs and performance issues, and also tidied things up in some of the tasks.
If you are using a previous version we strongly recommend that you upgrade.
In brief -
Improved processing for Pilatus images (both 6M and 2M)
Faster processing for large datasets in iMosflm
More reliability in indexing, refinement & integration
Better feedback
Easier multi-crystal strategy
Available from our website -
http://www.mrc-lmb.cam.ac.uk/harry/mosflm
======================================
In more detail (see http://www.mrc-lmb.cam.ac.uk/harry/mosflm/ver708/Release_notes for more) -
Mosflm
======
Non-Bragg spots (eg zingers and hot pixels) and ice spots are now
automatically removed when forming the standard profiles.
Improvements in mosaicity estimation and refinement.
Improvements in standard deviation estimates.
Improved spot finding parameters for in-house diffraction images
and for the very small spots that can be obtained from room
temperature (in situ screening) crystals.
More robust integration of images with very small spot separation.
iMosflm
======
The colour of the predicted reflections can be changed to improve
visibility on weak diffraction images.
Improved selection of the correct indexing solution in cases of
pseudosymmetry.
Scaling and merging now performed with AIMLESS rather than SCALA.
Release Notes for iMosflm version 1.0.6
New features
============
The speed of the interface has been much improved when processing a
large number of images in the Integration pane. The largest speed-up
came from not refreshing the profile displays afresh with every new
image integrated.
A greatly improved multi-crystal strategy option is available that
makes it much simpler to calculate a data collection strategy when
multiple crystals or multiple segments from one crystal are required
(see below, and also see the new iMosflm tutorial for details).
Bad spots identified during processing and whose numbers are plotted
during integration can now be displayed in the Image display window.
New mosaicity estimation will produce successive plots up to 8 degrees
if required.
Space group validation has been improved wherever a symbol can also be
typed in editable, pull-down lists (Indexing and Strategy panes).
The multi-sector, pull-down sectors list used in the Integration pane
has been adopted in the Indexing and Cell refinement panes.
Initial detector and crystal parameters are saved before refinement
and checked that they have not refined to unreasonable values. Users
may accept the warnings (& reset the parameters) or ignore them.
New items added to the Advanced integration tab of Processing options
includes provision for outliers affected by ice rings.
The maximum number of images in an integration block has been
increased from 20 to 200.
The FindHKL function has been placed within the Image display
window's toolbar.
Smaller rotation increments have been added to the Rotation: setting
in the Auto-complete menu in the Strategy pane.
The editable 'pie' widget displayed at the bottom of the Strategy pane
for any sector can now be adjusted in single degree steps (previously
only 5 degree changes were possible).
Individual images can now be deleted from the Images pane by selecting
them (left-click) and then right-clicking on the selected image. This
was previously only available for complete sectors.
Scala has been replaced by Aimless as the default scaling program used
by the QuickScale command button. Scala can be chosen from the
"Processing options->Advanced integration" tab under the Pointless &
Aimless/Scala switches.
Files containing lists of spots can be read into iMosflm and displayed
on the appropriate image. Files can be read from the main Session menu
item 'Read spots file...' and from the 'Read spots file...' button
(with an open file icon) next to the 'Index' button in the
Autoindexing pane.
Harry
--
Dr Harry Powell, MRC Laboratory of Molecular Biology, MRC Centre, Hills Road, Cambridge, CB2 0QH
----------
From: Jacob Keller
Where can one find a discussion of the differences between Aimless and Scala?
----------
From: Phil Evans
Aimless is a complete rewrite of Scala, but does essentially the same task. I haven't yet written it up properly, but there is a program document at
ftp://ftp.mrc-lmb.cam.ac.uk/pub/pre/aimless.html
A rough flow of the program is as follows:-
Read file from eg POINTLESS, sort data if necessary (SORTMTZ no longer needed)
Initial scale estimates from average intensities
First round scaling with small selection of strong reflections, chosen on I/sigI
First outlier rejection
Optimise the combination of profile-fitted (for weak spots) or summation integration intensities (for strong spots) from MOSFLM
First optimisation of sig(I) estimates
Main scaling on relatively strong reflections, chosen on normalised intensity E^2 (eg 0.8 < E^2 < 5)
Second outlier rejection
Final optimisation of sig(I) estimates
Final outlier rejection
Final statistics
Output of merged or unmerged data (MTZ or Scalepack format, or both)
Phil
Date: 16 April 2012 14:28
Dear all
We are pleased to announce the public release of new versions of iMosflm and Mosflm. We have addressed many bugs and performance issues, and also tidied things up in some of the tasks.
If you are using a previous version we strongly recommend that you upgrade.
In brief -
Improved processing for Pilatus images (both 6M and 2M)
Faster processing for large datasets in iMosflm
More reliability in indexing, refinement & integration
Better feedback
Easier multi-crystal strategy
Available from our website -
http://www.mrc-lmb.cam.ac.uk/harry/mosflm
======================================
In more detail (see http://www.mrc-lmb.cam.ac.uk/harry/mosflm/ver708/Release_notes for more) -
Mosflm
======
Non-Bragg spots (eg zingers and hot pixels) and ice spots are now
automatically removed when forming the standard profiles.
Improvements in mosaicity estimation and refinement.
Improvements in standard deviation estimates.
Improved spot finding parameters for in-house diffraction images
and for the very small spots that can be obtained from room
temperature (in situ screening) crystals.
More robust integration of images with very small spot separation.
iMosflm
======
The colour of the predicted reflections can be changed to improve
visibility on weak diffraction images.
Improved selection of the correct indexing solution in cases of
pseudosymmetry.
Scaling and merging now performed with AIMLESS rather than SCALA.
Release Notes for iMosflm version 1.0.6
New features
============
The speed of the interface has been much improved when processing a
large number of images in the Integration pane. The largest speed-up
came from not refreshing the profile displays afresh with every new
image integrated.
A greatly improved multi-crystal strategy option is available that
makes it much simpler to calculate a data collection strategy when
multiple crystals or multiple segments from one crystal are required
(see below, and also see the new iMosflm tutorial for details).
Bad spots identified during processing and whose numbers are plotted
during integration can now be displayed in the Image display window.
New mosaicity estimation will produce successive plots up to 8 degrees
if required.
Space group validation has been improved wherever a symbol can also be
typed in editable, pull-down lists (Indexing and Strategy panes).
The multi-sector, pull-down sectors list used in the Integration pane
has been adopted in the Indexing and Cell refinement panes.
Initial detector and crystal parameters are saved before refinement
and checked that they have not refined to unreasonable values. Users
may accept the warnings (& reset the parameters) or ignore them.
New items added to the Advanced integration tab of Processing options
includes provision for outliers affected by ice rings.
The maximum number of images in an integration block has been
increased from 20 to 200.
The FindHKL function has been placed within the Image display
window's toolbar.
Smaller rotation increments have been added to the Rotation: setting
in the Auto-complete menu in the Strategy pane.
The editable 'pie' widget displayed at the bottom of the Strategy pane
for any sector can now be adjusted in single degree steps (previously
only 5 degree changes were possible).
Individual images can now be deleted from the Images pane by selecting
them (left-click) and then right-clicking on the selected image. This
was previously only available for complete sectors.
Scala has been replaced by Aimless as the default scaling program used
by the QuickScale command button. Scala can be chosen from the
"Processing options->Advanced integration" tab under the Pointless &
Aimless/Scala switches.
Files containing lists of spots can be read into iMosflm and displayed
on the appropriate image. Files can be read from the main Session menu
item 'Read spots file...' and from the 'Read spots file...' button
(with an open file icon) next to the 'Index' button in the
Autoindexing pane.
Harry
--
Dr Harry Powell, MRC Laboratory of Molecular Biology, MRC Centre, Hills Road, Cambridge, CB2 0QH
----------
From: Jacob Keller
Where can one find a discussion of the differences between Aimless and Scala?
JPK
--
----------
From: Phil Evans
Aimless is a complete rewrite of Scala, but does essentially the same task. I haven't yet written it up properly, but there is a program document at
ftp://ftp.mrc-lmb.cam.ac.uk/pub/pre/aimless.html
A rough flow of the program is as follows:-
Read file from eg POINTLESS, sort data if necessary (SORTMTZ no longer needed)
Initial scale estimates from average intensities
First round scaling with small selection of strong reflections, chosen on I/sigI
First outlier rejection
Optimise the combination of profile-fitted (for weak spots) or summation integration intensities (for strong spots) from MOSFLM
First optimisation of sig(I) estimates
Main scaling on relatively strong reflections, chosen on normalised intensity E^2 (eg 0.8 < E^2 < 5)
Second outlier rejection
Final optimisation of sig(I) estimates
Final outlier rejection
Final statistics
Output of merged or unmerged data (MTZ or Scalepack format, or both)
Phil
No comments:
Post a Comment