Hi Gabor,
RELION will search within a specified (in class2d/3d) search range
during every iteration (this search range is determined automatically in
the 3D auto-refine). However, the center of the search will change from
one iteration to the next, such that particles can move much further
during multiple iterations. To prevent this, one can use a fixed, tight
standard deviation on the prior of the translations. You can specify the
standard deviation of this Gaussian prior (in pixels) using " -- offset
". And you can then keep that value fixed using " --fix_sigma_offset ".
By default, the prior is centered at (0,0): the center of the image. You
can change this for any value you want for every particle by adding two
columns to the input data.star file: rlnOriginXPrior and
rlnOriginYPrior. For example, you could set these values to the refined
values of rlnOriginX and rlnOriginY before you start your masked refinement.
Hope That Helps,
Sjors
On 02/05/2014 09:16 AM, Gabor PAPAI wrote:
> Hi Sjors,
>
> I would like to define the maximum allowed shift range
> in relion_refine. I tried to do 3D classification with a smaller mask
> than the particle - to see whether a factor was present. Unfortunately
> during the iterations Relion shifted the images quite substantially.
> Do you have an idea how to circumvent this?
>
> Thanks in advance!
>
> Best,
>
> Gabor
> ---
> Gabor Papai
> IGBMC
> Integrated Structural Biology
> 1, rue Laurent Fries, Bp 10142
> 67404 Illkirch, France
> Tel: +33-3-88655748
>
--
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
|