Hi,
> So finally to the question - in the particles.star file for the selected
> subset, is there anything I can use to tell me that two particles are
> derived from the same original ‘parent’ particle ?
Symmetry expansion preserves the particle image name (rlnImageName),
unless you also perform signal subtraction after it.
> I am able to Python
> hack something to do the analysis (which will take for ever as the
> particle count is in six figures), but I cannot see any obvious field in
> the particle.star file that carries the information I require. Distance
> between particle centres could work, but is going to be slow.
That should not be so slow. Perhaps you are running an O(N^2) algorithm,
where N is the number of total particles?
First group particles by micrograph (rlnMicrographName).
Then you should have ~ O(N^2 / M), where M is the number of micrographs.
You may further accelerate it by using nearest neighbour libraries but
probably it is not necessary.
Best regards,
Takanori Nakane
On 2019/10/07 9:18, Laurence Pearl wrote:
> We have a complex based on a substantial C3 symmetric core, which an
> additional pendant sub-complex which may or may not not fully reflect
> the symmetry of the core. Refining with the core symmetry applied gives
> poor resolution/fuzzy density for the sub-complex, whereas refining in
> C1 gives one good quality sub-complex, an additional symmetry related
> sub-complex with weaker density, and just the ghost of a third copy of
> the sub-complex. From this result we cannot distinguish whether we
> genuinely have a mixture of particles in which 1, 2 or 3 sub-complexes
> are bound, or wether all the particles have a single sub-complex, but
> some particles are misaligned by +/-120 degrees due to the string C3
> symmetry of the core.
>
> Either way by symmetry expanding the particles and reclassifying using a
> C3 asymmetric unit mask I can get distinct models for the occupied and
> unoccupied equivalent positions, and the subset of symmetry expanded
> particles that give the occupied asymmetric unit now refines well to a
> resolution where we can see the secondary structure - so problem solved.
>
> However, it would be very useful to know how whether there are in fact
> particles in that subset that are presenter more than once in different
> rotational settings, reflecting multiply loading of the sub-complex on
> the core, or whether the apparent multiple loading in the C1
> reconstruction was entirely due to misalignment.
>
> So finally to the question - in the particles.star file for the selected
> subset, is there anything I can use to tell me that two particles are
> derived from the same original ‘parent’ particle ? I am able to Python
> hack something to do the analysis (which will take for ever as the
> particle count is in six figures), but I cannot see any obvious field in
> the particle.star file that carries the information I require. Distance
> between particle centres could work, but is going to be slow.
>
> Laurence
>
> ----------------------------------------------------
> Laurence H. Pearl PhD FRS FMedSci
>
> Professor of Structural Biology
> Genome Damage and Stability Centre
> University of Sussex, Brighton, BN1 9RQ, UK
>
> Phone +44-(0)1273 678 349
> ----------------------------------------------------
> "Live Modestly and do Serious Things .. "
> - Dorothy Crowfoot Hodgkin
> ----------------------------------------------------
>
>
>
>
>
> ------------------------------------------------------------------------
>
> To unsubscribe from the CCPEM list, click the following link:
> https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCPEM&A=1
>
########################################################################
To unsubscribe from the CCPEM list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCPEM&A=1
|