Hello,
I am currently trying to classify virus-like particles in 2D. The sample is inherently heterogeneous and contains at least 4 different symmetric structures plus odd-shaped or incomplete particles.
I performed an initial 2D classification with an input diameter corresponding to the largest particles. Then I perform 2D classifications of same-sized particles with the appropriate diameter.
About 30 000 of 70 000 particles classify into a single class after a second 2D classification, with an estimated resolution at Nyquist (3.6A). This class is clearly not right though. There is a lot of heterogeneity if the particles are inspected nidividually, and there are no clear features in the protein shell (see attached image).
I tried to limit the resolution in 2D classification to 12 A, but the 2D classification still goes down to 5 A after 15 rounds already.
The command was:
--o Class2D/job104/run --i Select/job097/particles.star --dont_combine_weights_via_disc --scratch_dir Scratch --pool 6 --pad 2 --ctf --ctf_phase_flipped --iter 15 --tau2_fudge 2 --particle_diameter 330 --K 100 --flatten_solvent --zero_mask --strict_highres_exp 12 --oversampling 1 --psi_step 12 --offset_range 5 --offset_step 2 --norm --scale --j 3 --gpu ""
The output of the last iterations is as follows:
=============================
Expectation iteration 14 of 15
2.78/2.78 min ............................................................~~(,_,">
Maximization ...
2/ 2 sec ............................................................~~(,_,">
Estimating accuracies in the orientational assignment ...
0/ 0 sec ............................................................~~(,_,">
Auto-refine: Estimated accuracy angles= 2.6 degrees; offsets= 0.32 pixels
CurrentResolution= 5.14286 Angstroms, which requires orientationSampling of at least 1.78218 degrees for a particle of diameter 330 Angstroms
Oversampling= 0 NrHiddenVariableSamplingPoints= 67200
OrientationalSampling= 11.25 NrOrientations= 32
TranslationalSampling= 2 NrTranslations= 21
=============================
Oversampling= 1 NrHiddenVariableSamplingPoints= 2150400
OrientationalSampling= 5.625 NrOrientations= 256
TranslationalSampling= 1 NrTranslations= 84
=============================
Expectation iteration 15 of 15
3.07/3.07 min ............................................................~~(,_,">
Maximization ...
2/ 2 sec ............................................................~~(,_,">
__________________________________
What am I doing wrong? What else can I try to resolve the remaining heterogeneity? Playing with the T regularization parameter was unsuccessful. Would a mask, rather than just providing the particle diameter, be of any help?
Thanks,
Stephan
########################################################################
To unsubscribe from the CCPEM list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCPEM&A=1
|