Dear all,
Perhaps this is a good moment to highlight the link sent by Alan Brown
below. It encompasses almost 2 years of experience in building and
refining atomic models in high-resolution EM maps using COOT and REFMAC.
At LMB, these scripts and instructions have become an important part of
our structure determination pipeline. It may also be useful in your work.
Best,
Sjors
On 05/20/2015 09:30 AM, Alan Brown wrote:
> Hi Chris,
>
> We have a script (locres.csh), available from http://www2.mrc-lmb.cam.ac.uk/groups/murshudov/content/em_fitting/em_fitting.html that will calculate the local FSC for each chain in a folder of PDBs using Relion for both the mask creation and the post-processing.
>
> Best wishes,
>
> Alan
>
>
>
> On 20 May 2015, at 08:36, Sjors Scheres <[log in to unmask]> wrote:
>
>> Hi Chris,
>>
>> a. this is true: the 2 unfil.mrc will NOT be masked
>> b. you'll calculate a final FSC in the postprocessing program. You can
>> give any mask you like, and the randomized phase approach (see Chen et al
>> Ultramicroscopy 2013) will correct for its effects on the FSC curve. If
>> you let the mask cover only the best ordered bit of the map, it will be
>> higher than when you include the more floppy bits. This gives an
>> indication of local resolution, much like ResMap would. In a recenty
>> accepted paper on the spliceosome tri-snRNP we did this for many small
>> regions, and got very similar results as resmap.
>> Make sure your masks are always with a soft-edge, so their effects on the
>> FSC are not very large. It doesn't matter which program you use to make
>> the masks.
>> HTH,
>> Sjors
>>
>>
>>> Sjors and users-
>>>
>>> I am using a mask during refinements in Relion, which eliminates the small
>>> ribosomal subunit, thus, keeping its multiple conformations from affecting
>>> alignments driven by the lsu.
>>> I created this mask from a volume "made" in Chimera and then used
>>> relion_mask_create.
>>>
>>> My question is two-fold:
>>>
>>> a. The 3d half volumes made at each itr are masked by this volume, but I
>>> presume that the final 3d volume
>>> that is made when Relion converges will not be masked by this volume, is
>>> this correct?
>>>
>>> b. Given that part a above is true, then I presume that the final fsc
>>> calculated automatically in Relion will reflect the
>>> properties of the entire 3D volume, including the region that had been
>>> masked out during
>>> refinement, which may lower the overall resolution.
>>>
>>> While I can think of a number of ways to make a mask, what is the
>>> appropriate or best way to make
>>> a mask and the use Relion to calculate an fsc for the good region of the
>>> map.
>>>
>>> Of course ResMap can be used as a guide, but getting an fsc for the good
>>> region would be useful and
>>> in my reading its not clear how this is done.
>>>
>>> cheers,
>>>
>>> C Akey
>>>
>>
>> --
>> 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
|