If you find a bug or have a suggestion not on the list, please
let
me know .
EMAN1.9 is a bugfix release, intended as a final release of EMAN1 as it is gradually deprecated in favor of EMAN2. There were no 'major' changes in this release, strictly a large number of bugfixes and small improvements made over the last 2 years. The fixes are substantial enough that EMAN1.8 users should definitely upgrade to this version.s
Substantial improvements in refine2d.py. Strongly suggest using refine2d.py on all new data sets before moving on to 3D reconstruction.
Some programs use the EMAN2 style arguments ie - 'program <file> --option=value' rather than the old style 'program <file> option=value'
New HDF5 format compatible with EMAN2. Data can be stored in HDF5 files in many different ways. EMAN 1.9 will read both old (EMAN 1.7) HDF5 format files as well as new EMAN2 style HDF5 files. HDF5 writes are always in 'new' HDF5 format. Note that the current HDF I/O methods have poor performance on files with large numbers (~10,000 or more) of images in a single file. We're working on it...
New AIRS programs, such as 'skeleton'
Initial support for tetrahedral/cubic symmetry (thx to David Woolford and Phil Baldwin). Not well tested yet.
New options to make refinement work better on large icosahedral objects.
Improvements in parallelism infrastructure, though network-related problems may still exist on some linux clusters (please report problems)
Fixed random model generation in makeinitialmodel.py
Some linux clusters seem to have ssh-related problems. The symptom is either the refine command unreproducibly 'freezing' during the 'classalignall' process. If this happens, you can run 'killall ssh' on the node where refine is running, and the job will resume, though a few particles may be lost from the current reconstruction. The other problem is that you may observe some small fraction of the particles being missing from each iteration of the reconstruction. Both problems are relatively harmless, though annoying. Please report specific problems so we can try to establish what the unifying factor is...
dfilt option in refine works now. This is the major cause for the 1.7 release. This option was used for our recent 6 A reconstruction of GroEL, and we have been touting it as the best current approach for doing EMAN reconstructions (which it is). Unfortunately, due to an unfortunate error on my part, a critical change never got updated in the EMAN source repository. So, essentially, this option worked very well for me, but not for anyone else. Anyone who was experimenting with the dfilt option should give it another try. On the bright side, the bug wasn't causing any harm, specifying dfilt in the buggy version was almost equivalent to not specifying it at all. I apologize for this nasty snafu.
Better support for icosahedral reconstructions and for reconstructions of large models (contact wjiang@bcm.tmc.edu for more info)
Some programs didn't like EMAN's MRC files. They should now be properly CCP4 compliant.
EM file format support was minimal in last version, now properly handles stack files and byte-ordering
New program qplot utility for plotting 2 column text files.
EMAN can now compile and run on Macs (again). OS 10.3 only, with appropriate fink packages installed. This is still not officially supported.
The >2G file support was apparently broken in v1.5. This has been fixed.
Improved 64-bit support for use on Athlon-64's and Opterons. This is currently the fastest platform available for EMAN refinements (and the most cost-effective)
Support for KDE 3.2/Mandrake 10.0
Fixed funny narrow display in some cases in asymmetric triangle viewer
fitctf improved to work on a wider range of images, including particles from CCD frames
usefilt support in multirefine
MRC file i/o improved for use with chimera
New refinement options, phasecls, fscls, dfilt and refine for much more accurate image classification
Prototype animation module for Chimera (undocumented)
Bugfixes/improvements to PNG & PGM writing
MRC output now more comapatible with Chimera scaling
Fixed major alignment bug when box-size was not divisible by 4
Realfilters 110 and 111 for iterative expansion of masks
More binary releases
Substantial improvements were made to EMAN's parallelism. When used on small particles on a large number of processors, the previous versions of EMAN would run at less than 100% capacity on each processor. The new version should run at 100% on fairly large numbers of CPUs.
Python support - EMAN now officially supports Python scripting. Some very powerful operations can be performed with some very simple (and easy) scripts. Just to give you the flavor, here's a quick example which will compare class-averages to projections:
qindex - a program for very basic indexing of 2D crystal images was added. This program may be enhanced in future, depending on community response.
fitctf - improvements in fitting quality and accuracy. May still have problems with some CCD data.
procpdb.py - a script for basic manipulation of PDB files, designed largely for use with foldhunter
foldhunter - new option added to generate rotated/translated PDB files corresponding to the best fits
multirefine - now permits a different symmetry to be specified for each model. This is still an experimental option.
GNU configure -> CMake
Major documentation updates
Support for reading 8 and 16 bit tiffs (not tiled)
Support for image files/stacks larger than 2GB
Fixed some automatic boxing problems from (42)
Gnu 'configure' support for building EMAN
Better support for boxing helical 'filaments'
Many minor bugfixes
This page wasn't updated for over 2 years. All major
changes since compile (3) will be listed here now. Most of these
happened long ago. Hopefully I'll keep this up to date now
Changes
Gatan DM3 file format supported (read only)
Preliminary support for calling the EMAN library from Python using BOOST python.
Wider variety of 'standard' image formats (GIF, BMP, etc.) are supported now through the use of ImageMagick. NETPBM is no longer used.
'seteulers' command for reading and writing Euler angles to/from the image headers in various conventions.
New program 'qsegment' allows for interactive segmentation of 3D volumes. Very fast, but currently lacks isosurface display.
New SVD/MSA based refinement techniques embodied in svdrefine, svdmultirefine. Still quite experimental. Look for an upcoming publication.
proc2d no longer normalizes by default. Must be explicitly requested with 'norm', 'edgenorm', etc.
New MSA program 'svdcmp'
Bugfix in classalign2, now B-factors are only applied once
'meanshrink' option applies average filter rather than median in proc2d
'euler2=' was broken. Now it works again
Program 'raw2mrc' for conversion of raw binary data into an EM format.
New GUI for memories in CTFIT. Preparing for database interface.
Boxing of helices as a sequence of particles now supported in boxer. Will be improved in future.
Major bugfix in align3dsym
Focal pair merging program 'focalpair' and experimental mods so merged particles can be used in a reconstruction.
Resolution definition changed in pdb2mrc. Now uses standard EM definition (1/2 width of Gaussian in Fourier space)
'slow' option added in refine. Uses exhaustive search in alignment. Very slow, experimental.
'Detach' button in EMAN browser allows large images to be viewed.
Probe option in image control panel allows pixel values in an image to be probed, with a histogram display.
QHelp command provides simple documentation browser.
Multirefine works with fileserver for cluster parallelism
New normalization used in image comparison, now calculates proper normalized variance.
'selfcl' option in proc2d
'usefilt' option in refine allows arbitrarily filtered particles to be used for orientation/comparison, while the real data is still used for reconstruction.
'start.exclude' file allows specific particles to be excluded from reconstruction.
'ptcltrace' command allows analysis of self-consistency of particle Euler assignments over several refinement iterations
'alignhelix' command does primitive 2D alignments of helical 'particles'
Fileserver added to 'runpar'. This allows EMAN to be used on Linux clusters. Bypasses NFS for file writes as a workaround for several bugs in NFS. Support must be COMPILED in.
Tomographic in-plane alignment program.
Spider file writing fixed.
'glmatrix' and associated changes to provided dimensional reduction capabilities.
Particle tracking (through refinement iterations) added to several programs. Stored in 'particle.log'.
'reproj' option in project3d allows for quantitative comparison of projections and class-averages
'make3dreal' provides simple filtered back-projection based reconstruction.
'lst' filetype added. Simple text files that reference images in other files. Now used for 'cls' files, etc.
'align3dvec' a 3d alignment program using vectorization of volumes
better support for 'D' and 'oct' symmetries
Automasking in proc3d and refine (refine uses automask2= in proc3d)
'segment3d' k-means based automatic volume segmentation
Improved equation editing in ctfit
CTF correction now incorporates Wiener filter based on 1D structure factor of protein.
Changes
Documentation updated, much more complete.
New 'tree' option produces 2-5x speedup in refinements.
Redid parallelism, now supports workstation clusters as well as shared/distributed memory supercomputers.
New '3dit' and '3dit2' options for better 3D models from class averages.
New ctfit interface with database features for tracking changes in microscope/sample quality.
Median filters in boxer to aid in locating particles.
Fixed bug in t-test, previously used 1/4 data rather than 1/2.
Added print button to image windows.
Fixed masking/padding problems so edgemean is used rather than mean.
Fixed bug requiring initial drag in astigmatism slider on some SGIs.
New filefilt option in proc3d allows generic filtering.
Fixed bug in r/w certain MRC files.
Fixed display of angles in Euler view in eman. Selected point in asym tri was updated incorrectly.
Inconsistencies in step 1 and step 3 in use of CTF correction.
Classalign2 used to require median option when CTF correction was disabled. No longer.
Basic write support for VTK file format.
Known deficiencies/planned improvements
This is a BETA release. That is, it still has a few bugs, and there are still a few holes in the documentation. If you find a bug, PLEASE report it. If you're looking for missing documention, let me know. Requested items receive higher priority.
Optional Wiener filtration in CTF correction
Structure factors from PDB for Wiener filtration and CTF parameter determination.
Image drag/drop not finished yet.
More lowpass filter types neeeded.
Spider file support doesn't handel Euler angle conversion yet.
Support for 12 bit tiffs.
Changes
A serious bug in automatic boxing has been fixed
New option in boxer allows automatic box alignment for focal pairs
Fixed single image display in ctfit
Switched from QT1.44 -> QT2.0.1
Easier directory navigation in 'eman' file browser
Middle mouse button on plots brings up an inspector with picture saving
Changes
Many bug fixes in boxer.
particle view is now scalable
clicking on a particle to center it in the image view now works when the image view is scaled
Autoboxing thresholds work properly
When the particle view is more than 32768 pixels high, the display works now
All programs can now read Spider files, proc2d can write Spider files
Support for 'stack' files seems ok, single images may or may not work
Euler angles are not read properly from the Spider file headers yet.
When writing Spider files, all files will be 'stacks'
New option in refine: sep= allows particles to be put in multiple classes. This can make a BIG improvement, and only takes about 5-10% longer to run.
New 'Analyze' menu in eman/
Initial version of multirefine for functional state separation.
Multiple image view control panel (middle click) has many new options. drag/drop, delete, save file, etc.
cls files now have 4 digits, ie. cls012 -> cls0012
misc bugfixes
boxer crashes if you run it without specifying an image file to box.
Documentation is in a sorry state. Some links are missing.
Can only read single image Spider files. Stacks are broken.
The tutorial doesn't use the user parameters as much as it should.
'D' symmetry support isn't perfect yet.
Bug with reading 3D models in ctfit. Power spect. looks wrong.
Last Modified: