|Deletions are marked like this.||Additions are marked like this.|
|Line 1:||Line 1:|
|= Obsoleted, replaced by sxisa2.py =||= Obsoleted, replaced by sxisac2.py =|
Obsoleted, replaced by sxisac2.py
sxisac - 2D Clustering with ISAC: Iterative Stable Alignment and Clustering (ISAC) of a 2-D image stack.
usage in command line
sxisac.py stack_file output_directory --radius=particle_radius --img_per_grp=img_per_grp --CTF --restart_section --target_radius=target_radius --target_nx=target_nx --ir=ir --rs=rs --xr=xr --yr=yr --ts=ts --maxit=maxit --center_method=center_method --dst=dst --FL=FL --FH=FH --FF=FF --init_iter=init_iter --main_iter=main_iter --iter_reali=iter_reali --match_first=match_first --max_round=max_round --match_second=match_second --stab_ali=stab_ali --thld_err=thld_err --indep_run=indep_run --thld_grp=thld_grp --n_generations=n_generations --rand_seed=rand_seed --new --debug --use_latest_master_directory --stop_after_candidates --skip_prealignment
sxisac exists only in MPI version.
mpirun -np 176 --host <host list> sxisac.py bdb:data fisac1 --radius=120 --CTF > 1fou &
mpirun -np 176 --host <host list> sxisac.py bdb:data fisac1 --radius=120 --CTF --restart_section=candidate_class_averages,4 --stop_after_candidates > 1fou &
Note ISAC will change the size of input data such that they fit into box size 76x76 (see Description below).
The ISAC program needs MPI environment to work properly. Number of used MPI processes MUST BE multiplicity of given indep_run parameter (see parameters list below).
Depending on the cluster you are running, the way of using MPI will be significantly different. On some clusters,
- mpirun -np 32 sxisac.py ...
will be sufficient. On some clusters, one need to specify the host name:
- mpirun -np 32 --host node1,node2,node3,node4 sxisac.py ...
On some clusters, one need to submit a script to run MPI, please ask your system manager about how to run MPI program on your machine.
Also, different systems have different ways of storing the printout. On some clusters, printout is automatically saved. If it is not, it is recommended to use the linux command nohup to run the program, so that the printout is automatically saved to a textfile nohup.out. For example:
- nohup mpirun -np 32 sxisac.py bdb:test --img_per_grp=150 --generation=1
If there is no nohup on your system, it is recommended to redirect the printout to a textfile.
mpirun -np 32 sxisac.py bdb:test --img_per_grp=150 --generation=1 > output.txt
To restart a run that stopped intentionally or unintentionally use the '--restart_section' option.
2-D images in a stack file (bdb or hdf): images have to be square (nx=ny) (default required string)
- particle radius: there is no default, a sensible number has to be provided, units - pixels (default required int)
- number of images per class: in the ideal case (essentially maximum size of class) (default 100)
- apply phase-flip for CTF correction: if set the data will be phase-flipped using CTF information included in image headers (default False)
- restart section: each generation (iteration) contains three sections: 'restart', 'candidate_class_averages', and 'reproducible_class_averages'. To restart from a particular step, for example, generation 4 and section 'candidate_class_averages' the following option is needed: '--restart_section=candidate_class_averages,4'. The option requires no white space before or after the comma. The default behavior is to restart execution from where it stopped intentionally or unintentionally. For default restart, it is assumed that the name of the directory is provided as argument. Alternatively, the '--use_latest_master_directory' option can be used. (default ' ')
- target particle radius: actual particle radius on which isac will process data. Images will be shrinked/enlarged to achieve this radius (default 29)
target particle image size: actual image size on which isac will process data. Images will be shrinked/enlarged according to target particle radius and then cut/padded to achieve target_nx size. When xr > 0, the final image size for isac processing is 'target_nx + xr - 1' (default 76)
- x range: of translational search. By default, set by the program. (default 1)
- output directory name: into which the results will be written (if it does not exist, it will be created, if it does exist, the results will be written possibly overwriting previous results) (default required string)
For each generation of running the program, there are two phases. The first phase is an exploratory phase. In this phase, we set the criteria to be very loose and try to find as much candidate class averages as possible. This phase typically should have 10 to 20 rounds (set by max_round, default = 20). The candidate class averages are stored in class_averages_candidate_generation_n.hdf.
The second phase is where the actual class averages are generated, it typically have 3~9 iterations (set by match_second, default = 5) of matching. The first half of iterations are 2-way matching, the second half of iterations are 3-way matching, and the last iteration is 4-way matching. In the second phase, three files will be generated:
class_averages_generation_n.hdf : class averages generated in this generation, there are two attributes associated with each class average which are important. One is members, which stores the particle IDs that are assigned to this class average; the other is n_objects, which stores the number of particles that are assigned to this class average.
class_averages.hdf : class averages file that contains all class averages from all generations.
generation_n_accounted.txt : IDs of accounted particles in this generation
generation_n_unaccounted.txt : IDs of unaccounted particles in this generation
Time and Memory
Unfortunately, ISAC program is very time- and memory-consuming. For example, on my cluster, it takes 15 hours to process 50,000 64x64 particles on 256 processors. Therefore, before embarking on the big dataset, it is recommended to run a test dataset (about 2,000~5,000 particles) first to get a rough idea of timing. If the timing is beyond acceptable, the first parameter you may want to change is max_round, you may change it to 10 or even 5 with mild effects on the results.
Retrieval of images signed to selected group averages
- Open in e2display.py file class_averages.hdf located in the main directory.
- Delete averages whose member particles should not be included in the output.
- Save the selected subset under a new name,say select1.hdf
- Retrieve IDs of member particles and store them in a text file ohk.txt:
- sxprocess.py --isacselect class_averages.hdf ok.txt
- Create a vritual stack containng selected particles:
- e2bdb.py bdb:data --makevstack:bdb:select1 --list=ohk.txt
The same steps can be performed on files containing candidate class averages.
RCT information retrieval
Let us assume we would want to generate a RCT reconstruction using as a basis group number 12 from ISAC generation number 3. We have to do the following steps:
- Retrieve original image numbers in the selected ISAC group. The output is list3_12.txt, which will contain image numbers in the main stack (bdb:test) and thus of the tilted counterparts in the tilted stack. First, change directory to the subdirectory of the main run that contains results of the generation 3. Note bdb:../data is the file in the main output directory containing original (reduced size) particles.
- cd generation_0003
- sxprocess.py bdb:../data class_averages_generation_3.hdf list3_12.txt --isacgroup=12 --params=originalid
- Extract the identified images from the main stack (into subdirectory RCT, has to be created):
- e2bdb.py bdb:test --makevstack=bdb:RCT/group3_12 --list=list3_12.txt
- Extract the class average from the stack (NOTE the awkward numbering of the output file!).
- e2proc2d.py --split=12 --first=12 --last=12 class_averages_generation3.hdf group3_12.hdf
- Align particles using the corresponding class average from ISAC as a template (please adjust the parameters):
- sxali2d.py bdb:RCT/group3_12 None --ou=28 --xr=3 --ts=1 --maxit=1 --template=group3_12.12.hdf
- Extract the needed alignment parameters. The order is phi,sx,sy,mirror. sx and mirror are used to transfer to tilted images.
- sxheader.py group3_12.12.hdf --params=xform.align2d --export=params_group3_12.txt
The program will perform the following steps (to save computation time, in case of inadvertent termination, i.e. power failure or other causes, the program can be restarted from any saved step location, see options) :
- The images in the input stacked will be phase-flipped.
- The data stack will be pre-aligned (output is in subdirectory 2dalignment, in particular it contains the overall 2D average aqfinal.hdf, it is advisable to confirm it is correctly centered).
- In case 2dalignment directory exists steps 1 and 2 are skipped.
- The alignment shift parameters will be applied to the input data.
IMPORTANT: Input aligned images will be resized such that the original user-provided radius will be now target_radius and the box size target_nx + xr - 1. The pixel size of the modified data is thus original_pixel_size * original_radius_size / target_radius.
- The pseudo-code for adjusting the size of the radius and the size of the images is as follows:
- shrink_ratio = target_radius / original_radius_size
- new_pixel_size = original_pixel_size * shrink_ratio
- if shrink_ratio is different than 1: resample images using shrink_ratio
if new_pixel_size > target_nx : cut image to be target_nx in size
if new_pixel_size < target_nx : pad image to be target_nx in size
- The newly added options target_radius and target_nx allow the user to finely adjust the image so that it contains enough background information.
- The program will iterate through generations of ISAC by alternating two steps. The outcome of these two steps is in subdirectory generation_*** (stars replaced by the current generation number).
- Calculation of candidate class averages.
- saved checkpoint: restart from just before this step with --restart_section=candidate_class_averages,4 for the fourth isac generation.
- Calculation of validated class averages.
- saved checkpoint: restart from just before this step with --restart_section=reproducible_class_averages,4 for the fourth isac generation.
- Calculation of candidate class averages.
- The program will terminate when it cannot find any more reproducible class averages.
- If no restart option is given the program will pick-up from the last saved point.
See the reference below.
Yang, Z., Fang, J., Chittuluru, F., Asturias, F. and Penczek, P. A.: Iterative Stable Alignment and Clustering of 2D Transmission Electron Microscope Images, Structure 20, 237-247, February 8, 2012.
Author / Maintainer
Horatiu Voicu, Zhengfan Yang, Jia Fang, Francisco Asturias, and Pawel A. Penczek
- category 1
- works for author, works for others as far as the author knows.
None right now.