Hi Sjors,
Thank you for the quick response! We think there might be two issues that could be contributing to the artifact. First, our helical tube has some variation in diameter(~ 1 nm), that 3D classification was not able to separate. Second, the inner radial density of the tube is known to be more floppy than the more outer regions. We plan to classify by the diameter using a different program in the near future. Do you think this could help? Also, would it be worth masking out the inner radial density?
Thanks for your help!
Leo
________________________________________
From: Sjors Scheres [[log in to unmask]]
Sent: Wednesday, December 07, 2016 12:10 PM
To: Kong, Leopold (NIH/NIDDK) [F]; [log in to unmask]
Subject: Re: [ccpem] helical reconstruction and masking
Hi Leo,
The masking will not generate these artefacts. They are already present
in your refined map (albeit perhaps less visible prior to the
sharpening). It could be that the helical symmetry isn't correct, or it
could be that some floppy densities are sticking inside the helix. Does
the rest of the subunit show good 4.2A protein-like features, or is
everything a bit weird? The latter seems the case (from your pdf), which
may indicate helical symmetry isn't quite right yet. I also don't really
like the shape of your FSC curve.... But ultimately these things are
hard to tell from a distance.
HTH,
Sjors
On 12/07/2016 04:51 PM, Kong, Leopold (NIH/NIDDK) [F] wrote:
> Hi,
>
> I've been using helical processing in relion for a dataset. After 3D auto refine, I got a map (upper right corner, purple), that seems reasonable to be ~ 6 Angstrom. Relion estimated it to be ~ 4.4 Angstrom. I did post processing with a mask (right for FSC curves of pre and post processing), which resulted in the green map (lower right corner). Interestingly, there seems to be a lot of highly regular radially smeared density, which I suspect to be some kind of artifact from the masking.To generate the mask, I used the following parameters in relion: lowpass filter map of 25 A, pixel size 1.08 A (from data collection parameters), 0.001 initial binarisation threshold, extended binary map 2 pixels, and an 8 pixel soft-edge. I would greatly appreciate any advice on how to improve the post processing, or insight into whether something else might be going on to induce the potential artifacts.
>
> Thanks!
>
> Leo
>
>
>
>
--
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
|