Manpage of removeDuplicates
Section: User Commands (1)
Return to Main Contents
removeDuplicates - remove duplicate particles
removeDuplicates filename iterationNumber
removeDuplicates filename iterationNumber classID
removeDuplicates filename iterationNumber classID flgShuffle
removeDuplicates filename iterationNumber classID flgShuffle onlyTomNum
Alignment can result in multiple model particles pointing
to essentially the same location and orientation, especially
when the search range is comparable to or larger than the spacing
between particles. If flgRemoveDuplicates is set to 1 in
parameter file 'filename', removeDuplicates will identify
such duplicates in the motive list(s), and will remove
them from further consideration by assigning them to the duplicate
class (ID = -9999). Particle class is stored in column 20 of the
motive list, with a default of 0. If a particle is marked as a duplicate,
previous particle class will be preserved in column 16 of the motive list.
All particles are always restored to their previous class before
search for duplicates. RemoveDuplicates should be run immediately following
mergeEm. Programs averageAll, calcFSC, and calcSSNR
ignore members of the duplicate class unless -9999 has been explicitly
added to selectClassID.
Conflicts between 'duplicates' can be resolved in multiple ways.
RemoveDuplicates chooses a particular solution using hueristics
which try to minimize the number of particles removed while preserving
particles having high cross-correlation scores with the reference.
At iterations after the first, removeDuplicates restores the original
particle class before performing a fresh search for duplicates. This
allows particles which were removed as duplicates to potentially be
restored at later iterations, e.g. if duplicateShiftTolerances become
Normally, tomograms are processed independently except that multiple
tomograms refering to the same volume (e.g. when using duplication to
enforce symmetry) are processed together. The test for identity
currently relies on string comparison of the literal paths, however, so,
recognized as pointing to the same volume.
The file name of the parameter file. See PEET(1) man page
for descriptions of applicable parameter file settings.
A sample parameter file, with parameter descriptions,
is available at $PARTICLE/templates/master.prm.
RemoveDuplicates uses the following parameters:
- iterationNumber (optional)
An integer specifying the alignment iteration whose output is to be
processed. The name of the corresponding motive list file will
contain iterationNumber + 1, since it refers to the iteration
for which the file will serve as input.
- classID (optional)
An integer specifying the class ID to be used for duplicate particles.
(Default = -9999).
If non-zero, different initialization of the pseudo-random number
generator will be used at each invocation. (Default = 0).
If non-zero, duplicate removal, if enabled, will be done on multiple
volume table entries in parallel, yielding faster execution. This
option should be disabled if there are volume table entries with
unique names but which point to the same volume (e.g. via hard or
symbolic links). (Default = 0).
If greater than 0 specifies the number of the volume to process; if
omitted or less than or equal to 0, all tomograms will be processed
PEET(1), prepareEM(1), prepareRef(1), mergeEM(1), alignSubset(1), averageAll(1),
calcFSC(1), and calcSSNR(1)
- SEE ALSO
This document was created by
using the manual pages.
Time: 18:16:05 GMT, January 11, 2021