[3dem] model-map FSC @ 0.25 cutoff?
Sjors Scheres
scheres at mrc-lmb.cam.ac.uk
Fri May 4 08:33:41 PDT 2018
Hi Gabe,
I am afraid we may have at least contributed to this mess. When we
published ~6A maps of the spliceosome in 2015
(https://www.nature.com/articles/nature14548.pdf), in extended figure 4e
we drew a line at FSC=0.25 in the FSC model-vs-map plot. The reasoning
at the time was to relax the 0.5 criterion (which is normally used) to
account for the fact we did not refine our PDB models in the
moderate-resolution maps (at 6A, refinement is most likely to result in
overfitting; instead we just rigid-body placed crystal structures and
homology models). If I remember correctly, we showed these 'resolutions'
in response to a reviewer request, but in hindsight this was a mistake,
as it may have led to more people using this criterion in a range of
situations. As far as I know, there is no reasoning behind the FSC=0.25
criterion, as there is for the 0.143 criterion between noisy half-maps
and the 0.5 criterion for model-vs-map (see Rosenthal & Henderson, JMB,
2003). The FSC=0.25 criterion should probably not be used, and I wished
we hadn't done so in that paper.
HTH,
Sjors
On 05/04/2018 12:30 AM, Gabriel Lander wrote:
> Hi all,
> I've recently been seeing model-to-map FSCs being reported using a cutoff of 0.25 - I must have missed this discussion or the paper describing the rationale for this cutoff value instead of 0.5. Can someone fill me in?
> -gabe
>
> PS. I take full responsibility for all repercussions/damages stemming from asking an FSC-based question on this listserv.
> _______________________________________________
> 3dem mailing list
> 3dem at ncmir.ucsd.edu
> https://mail.ncmir.ucsd.edu/mailman/listinfo/3dem
--
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
More information about the 3dem
mailing list