Hi guys,
I'll look at this once I get back from a trip, next week. Several similar
reports of erroneous STAR files upon extraction in recent days may
indicate that there is a new bug somewhere. If so, my suspicion would be
that it could have to do with manually deleting or adding particles after
an autopicking run.
Meanwhile, when in doubt, one can check the number of columns in the
particles.star file with the following command:
grep mrc particles.star | awk '{print NF}' | sort | uniq
which should only give a single number. If it gives multiple numbers, then
something is corrupt! You could then find lines (with for example 33
columns) using:
grep mrc particles.star | awk '{if (NF==33) print}'
Hope that helps,
Sjors
Eric,
>
> As I explained, the --angpix flag does not matter if you are getting
> this error, because RELION has found sufficient information
> (magnification and detector pixel-size) in the .star-file to calculate
> it, instead of using any user-provided input. Either fixing or removing
> those data-fields in the .star-file is the only way to immediately fix
> the issue. That being said, it might stem from inside RELION (i.e.
> having produced a bad .star-file to begin with), so please attach you
> .star-file to the previously linked issue
> <https://bitbucket.org/tcblab/relion2-beta/issues/143/relion-cannot-read-particlesstar-after>,
> that will help us determine the root of the issue.
>
> Thanks!
>
> /Björn
>
>
> On 01/11/2017 08:06 AM, Eric Durand wrote:
>> Hi Björn,
>>
>> I do have the very same problem concerning the input of the pixel
>> size. Please, could you tell me where to put the --angpix flag
>> argument: into the particles.star file, or into the RELION GUI?
>>
>> Many thanks
>> Best regards
>> Eric
>>
>>
>>
>> Le Mardi 10 janvier 2017 16h21, Bjoern Forsberg
>> <[log in to unmask]> a écrit :
>>
>>
>> Hi Donna,
>>
>> The error is given as an indication that you have inconsistent
>> meta-datain your particle .star-file. You can specify the pixel-size
>> (in Angstroms) by using the --angpix flag, but if RELION finds that
>> the particle .star-file conatains both
>>
>> 1. Magnification
>> 2. Physical pixel size of the detector
>>
>> then it overrides the --angpix flag and calculates the pixel-size on
>> its own. In your case, the first particle in your file is calculated
>> to have 4.4 A/pixel, whereas at some point it finds a particle where
>> the calculated pixel-size is 5288660 A/pixel, which is clearly wrong.
>> I suspect that you or relion has missed some fields in the .star file
>> so that the wrong value is picked up and used for the pixel-size
>> calculation. If you find that your .star-file is consistent with this
>> issue, I'd appreciate if you could upload it there, to help us sort it
>> out.
>>
>> Thanks!
>>
>> /Björn
>>
>> On 01/10/2017 03:36 PM, Matzov Donna wrote:
>> Hi All,
>>
>> I have some 2D classification issues. After doing an initial 2D
>> classification I wasn't very pleased with it so I figured I should go
>> over the micrographs and clean them. After doing so I extracted again
>> the particles and tried to run 2D classification again but I got this
>> error message:
>>
>> "first_angpix= 4.4 my_angpix= 5.28866e+06 mag= 0.105887 dstep= 56
>> MlOptimiser::initialiseGeneral: ERROR inconsistent magnification and
>> detector pixel sizes in images in input STAR file
>> File: /opt/relion/relion2-beta/src/ml_optimiser.cpp line: 1234
>> 2 more processes have sent help message help-mpi-api.txt / mpi-abort
>> Set MCA parameter "orte_base_help_aggregate" to 0 to see all help /
>> error messages"
>>
>> I tried to go back to re-run the 2D class that originally worked but
>> got the same message.
>>
>> Any what is the issue and how to fix it?
>>
>> Thanks,
>>
>> Donna
>>
>>
>>
>>
>
>
--
Sjors Scheres
MRC Laboratory of Molecular Biology
Francis Crick Avenue, Cambridge Biomedical Campus
Cambridge CB2 0QH, U.K.
tel: +44 (0)1223 267061
http://www2.mrc-lmb.cam.ac.uk/groups/scheres
|