Dear Peter,
Some labels of the metadata entries change from one version to the next.
I am not sure what changed between 1.1 and 1.3 that is causing your
specific problem. Are you trying to continue an auto-refinement? You
could run a small 3D refinement in 1.3 to find out the differences: they
should be small and you could then re-edit the old 1.1 star files to
follow the 1.3 conventions.
Anyway, your question seems to indicate that you merely want to do
postprocessing. For that you only need the 2 unfil.mrc maps, so no STAR
files. However, I am not sure whether 1.1 already wrote those out
(possibly not). If you don't have the unfil.mrc files you will need to
re-run a refinement job. You should be able to do that in 1.3 much like
you did in 1.1.
If you could somehow retrieve the coordinates from your EMAN data base,
that would also help: you could then follow Ludo's suggestion.
HTH,
S
On 12/17/2014 01:30 PM, Ludovic Renault wrote:
> Oh, sorry I didn't get that bit right away,
>
> I have indeed use phase flipped stacks successfully in the past. spi format
> usually. I was using the relion_preprocess_mpi command that was generating
> the appropriate normalized mrcs and star files.
> from eman2 you can convert hed/img to spi files.
> mpirun -n 10 'which relion_preprocess_mpi` --o name --operate_on
> name_flip.spi --norm --bg_radius 70 --white_dust 6 --black_dust 6
> for example
> I just tried and you can still do that in 1.3 relion.
> Let me know if that what you want to do.
> Cheers.
> Ludo
>
> On Wed, Dec 17, 2014 at 1:13 PM, Peter Engelhardt <
> [log in to unmask]> wrote:
>> Hi Ludo,
>>
>> Many thanks Ludo for your suggestion that we have indeed been aware of 'as
>> a less suitable solution' as we never saved the 'save box files’ in EMAN2.
>> So, apparently, - if there are no other solutions and suggestions - it
>> seems we must 'start all over again' with “Getting higher resolution and
>> map sharpening” in Relion 1.3.
>>
>> Using EMAN2 'the phase-flipped' particles stacks, saved as IMAGIC (.hed
>> and .img) format as input files is in Relion 1.1 is most simple and
>> straight forward to get 3D class and 3D refined maps at high resolution
>> (including .star files).
>>
>> However, to continue in Relion 1.3, as it appears now, is not possible
>> unfortunately.
>> However, to inspect i.e. visualise in Relion 1.3, Relion 1.1 .star files
>> of 2D (e.g. 100 classes), works perfectly but to save e.g. all “good”
>> classes containing particles is possible but the ’saved’ (.star) files are
>> NOT ‘recognised’ i.e. working in Relion 1.3 for further processing.
>>
>> We still hope there would be 'an elegant' solution to this annoying
>> problem?
>>
>> Thank you in advance
>>
>> Cheers
>> Peter
>>
>>
>>
>>
>> On 16 Dec 2014, at 21:25, Ludovic Renault <[log in to unmask]> wrote:
>>
>> Hi Peter,
>> You can save box files from eman2 and use them directly with relion-1.3.
>> You specifically have to tick the 'save box files' into eman2 to get them.
>> Instead of "_autopick.star" use suffix ".box" at the extraction step in
>> relion.
>> Should work flawlessly.
>> Ludo
>>
>> On Tue, Dec 16, 2014 at 12:50 PM, Peter Engelhardt <
>> [log in to unmask]> wrote:
>>> Dear Sjörs,
>>>
>>> Problems in RELION that we would be most happy to get solutions to:
>>>
>>> - We used IMAGIC (.hed and .img) of particles stacks picked with EMAN2
>>> with RELION (1.1 version) with success.
>>> However, though the “Conventions & File formats” in RELION 1.3 version
>>> still seem to support IMAGIC stacks we have NOT been able to continue:
>>> “Getting higher resolution and map sharpening” as not only IMAGIC stacks
>>> - but also .star files produced in RELION 1.1 - do not seem to be
>>> supported by RELION 1.3.
>>> We want to use the EMAN2 picked particle stack to continue. We have also
>>> tested picking with RELION 1.3 but, in our case, the results are different.
>>> We want to also to compare EMAN2 with RELION 1.3 picked particles.
>>> So how to proceed?
>>>
>>> - MTF of our TEMs detectors are also needed for “Getting higher
>>> resolution and map sharpening” and we have sent Gatan requirements on this
>>> data - but not got any answers? Here the questions:
>>> "The MTF (modulation transfer function)-curves (also in .star formats) as
>>> indicated in
>>> http://www2.mrc-lmb.cam.ac.uk/relion/index.php/Analyse_results
>>> for GATAN, ULTRASCAN 4000, CCD 4x4k detectors for our TEMs:
>>> 1. JEOL JEM-2200FS
>>> http://nmc.aalto.fi/en/instruments/tem/jem-2200fs/
>>> 2. JEOL JEM-3200FSC
>>> http://nmc.aalto.fi/en/instruments/tem/jem-3200fsc/“
>>>
>>> We thank you in advance
>>>
>>> Cheers
>>> Peter
>>> _________________________________________________________
>>> Peter Engelhardt, PhD, docent
>>> Adjunct Professor in Molecular Genetics
>>> Unit of Molecular Electron Tomography
>>>
>>> Nanomicroscopy Center (NMC ), Department of Applied Physics
>>> Aalto University, School of Science and Technology,
>>> Puumiehenkuja 2, FI02150 Espoo
>>> Finland
>>>
>>> Department of Pathology
>>> Haartman Institute
>>> P.O.Box 21, FI-00014 Univ. of Helsinki, Finland
>>> Visiting Address: Haartmaninkatu 3, Helsinki
>>>
>>> Home address:
>>> Lindstedstvägen 1 B 7
>>> FI-02700 Grankulla,
>>> Finland
>>>
>>> Mobile: +358-(0)400193342
>>> Email: [log in to unmask], [log in to unmask],
>>> Peter Engelhardt<[log in to unmask]>
>>> URL: http://www.lce.hut.fi/~engelhar/
>>>
>>>
>>>
>> _________________________________________________________
>> Peter Engelhardt, PhD, docent
>> Adjunct Professor in Molecular Genetics
>> Unit of Molecular Electron Tomography
>>
>> Nanomicroscopy Center (NMC ), Department of Applied Physics
>> Aalto University, School of Science and Technology,
>> Puumiehenkuja 2, FI02150 Espoo
>> Finland
>>
>> Department of Pathology
>> Haartman Institute
>> P.O.Box 21, FI-00014 Univ. of Helsinki, Finland
>> Visiting Address: Haartmaninkatu 3, Helsinki
>>
>> Home address:
>> Lindstedstvägen 1 B 7
>> FI-02700 Grankulla,
>> Finland
>>
>> Mobile: +358-(0)400193342
>> Email: [log in to unmask], [log in to unmask],
>> Peter Engelhardt<[log in to unmask]>
>> URL: http://www.lce.hut.fi/~engelhar/
>>
>>
>>
--
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
|