Re: [scipion-users] movie alignments

[Carlos O. S. <co...@cn...>]

Dear all,


On 04/29/17 17:05, Grigory Sharov wrote:
> Hi Dmitry,
>
> I can try to answer a couple of your questions:
>
> I assume you are in release-1.1 branch (Scipion v1.1-beta (2017-04-21) 
> Balbino)?
>
> 1) I think it's a bad idea to use aligned movies as input for a second 
> alignment - the interpolation will cause you more problems and will 
> not improve any alignment. Also, motioncor2 cannot read input shifts 
> for the movies.
If I'm right, Scipion is not producing an interpolated output movie, 
unless the corresponding option is activated. The default behavior is to 
attach shift parameters to each frame, so that the next protocol can 
align the frames properly. I don't know the details, but I tend to agree 
with Grigori in that motioncor2 cannot read the shifts.

It makes sense to correct for global shifts (motioncorr or xmipp 
correlation) and then for local shifts with optical flow. As far as I 
understand motioncorr2 cannot produce aligned frames, and I would also 
be suspicious of the double interpolation (the one done by motioncorr2 
and the one by optical flow).

I find less useful a double correction by global shifts.

Kind regards, Carlos Oscar
>
> 3) hm, can't seem to reproduce your error (neither in normal, no in 
> streaming mode). Is it failing in the middle of the run (not first few 
> movies)?
>
> Best regards,
> Grigory
>
> --------------------------------------------------------------------------------
> Grigory Sharov, Ph.D.
>
> MRC Laboratory of Molecular Biology,
> Francis Crick Avenue,
> Cambridge Biomedical Campus,
> Cambridge CB2 0QH, UK.
> tel. +44 (0) 1223 267542 <tel:+44%201223%20267542>
> e-mail: gs...@mr... <mailto:gs...@mr...>
>
> On Sat, Apr 29, 2017 at 10:35 AM, Dmitry Semchonok 
> <sem...@gm... <mailto:sem...@gm...>> wrote:
>
>     Dear colleagues,
>
>     The issue is again with movie correction scripts.
>
>     1. Motioncorr2
>     As input, I use the movie stacks (can I use corrected movies as an
>     input to increase the quality - for example after
>     xmipp3-correlation alignment or it will be repetition?) and get
>     the error  -
>
>     ProgrammingError: Incorrect number of bindings supplied. The
>     current statement uses 32, and there are 22 supplied.
>     (the git was updated)
>
>     Встроенное изображение 1
>
>     How can I fix that?
>
>
>     2. xmipp3 - correlation alignment / optical alignment
>
>
>
>
>     2.1 correlation alignment
>
>     - from time to time the protocol stops. When I resubmit it
>     continues working.
>     But the question is why does that happen: problem with my dataset
>     and/or bug?
>
>
>     2.2 optical alignment
>
>     - if you may remember the problem was that it finished because of
>     an error in 1 file.
>
>     Is there already the solution for skipping the movies that causing
>     the error?
>
>     Alternatively, I removed that file (it looked fine in terms of
>     frames and visibility - I can send it for the analysis) and have
>     to rerun the xmipp3-correlation again. But there is no guarantee
>     that the program will not stop on some other file.
>
>
>
>     3. grigoriefflab - Input - movie stacks
>
>     stops in the beginning with error -
>
>     Input stack filename
>     14274:  
>     [Runs/022622_ProtUnblur/tmp/movie_000010/FoilHole_26594251_Data_25564934_25564935_20160826_2014-21929.mrc]
>     : 24
>     14275:
>     14276:    No Such File - please enter a valid filename
>
>
>
>
>     Встроенное изображение 2
>
>
>
>     Sincerely,
>     Dmitry
>
>
>     ------------------------------------------------------------------------------
>     Check out the vibrant tech community on one of the world's most
>     engaging tech sites, Slashdot.org! http://sdm.link/slashdot
>     _______________________________________________
>     scipion-users mailing list
>     sci...@li...
>     <mailto:sci...@li...>
>     https://lists.sourceforge.net/lists/listinfo/scipion-users
>     <https://lists.sourceforge.net/lists/listinfo/scipion-users>
>
>
>
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
>
>
> _______________________________________________
> scipion-users mailing list
> sci...@li...
> https://lists.sourceforge.net/lists/listinfo/scipion-users

-- 
------------------------------------------------------------------------
Carlos Oscar Sánchez Sorzano                  e-mail:   co...@cn...
Biocomputing unit                             http://biocomp.cnb.csic.es
National Center of Biotechnology (CSIC)
c/Darwin, 3
Campus Universidad Autónoma (Cantoblanco)     Tlf: 34-91-585 4510
28049 MADRID (SPAIN)                          Fax: 34-91-585 4506
------------------------------------------------------------------------