If it's related to the psi-prior: is your psi-angle sampling rate fine
enough compared to the prior? I would use a sampling rate that is at
least 3-4x smaller than the prior.
HTH,
Sjors
On 04/15/2016 10:52 AM, Juha Huiskonen wrote:
> Hi Sjors,
>
> Forgot to mention that the problem occurs also in 2D classification and
> it's definitely a problem with the PsiPrior. If I omit the --sigma_psi
> parameter, it runs fine (but I want to limit the in-plane rotation as I
> know it already pretty accurately).
>
> In my STAR file I have set AnglePsi and AnglePsiPrior to the same value.
> Even if I remove the AnglePsiPrior (or AnglePsi), the problem persists.
>
> I have checked the particle images visually and using relion_image_handles
> --stats. I couldn't spot anything suspicious.
>
> Any suggestions?
>
> As IDs are randomised, in a future release of Relion it would be useful to
> print out the name of the particle so that if can be removed. Many thanks.
>
> Best wishes,
> Juha
>
>
> On Thu, Apr 14, 2016 at 12:52 PM, Sjors Scheres <[log in to unmask]>
> wrote:
>
>> Hi Juha,
>> Yes, these things sometimes happen with odd particles. Does the sorting
>> reveal anything odd? Or a simple analysis of statistics?
>> The numbering gets randomised, I'm afraid.
>> S
>>
>> On 04/14/2016 09:31 AM, Juha Huiskonen wrote:
>>
>>> Dear Sjors and Relion users,
>>>
>>> I am encountering the following error during 3D classification:
>>>
>>> pdf_offset= 0 pdf_orientation= 0.0037792 [oo]
>>> ipart= 0 part_id= 2718
>>> iorient= 0 idir= 0 ipsi= 0
>>> exp_directions_prior[0]= 1
>>> exp_psi_prior[0]= 0.0037792
>>> ...
>>> exp_psi_prior[9]= 0.00601489
>>> ERROR! pdf_offset==0.|| pdf_orientation==0.
>>> File: src/ml_optimiser.cpp line: 4130
>>>
>>> It seems particle number 2718 is somehow bad. Should I simply remove the
>>> 2718th line from the star file? Does numbering start from 0 or 1? (I might
>>> be missing something as I tried both. As it didn't work I suspect the
>>> numbering is not sequential somehow.)
>>>
>>> Best wishes,
>>> Juha
>>>
>>>
>> --
>> 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
>>
>>
--
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
|