[phenixbb] real space correlation

[Pavel Afonine]

Hi Abhinav,

> When does it write CC per atom vs per residue? Can you please elaborate?

it's resolution dependent.

Say you are looking at ARG (or something similar with long side chain), 
which has a few atoms sticking out of density, while the other atoms are 
well placed into density. Then if you compute CC for the whole residue 
it will still be high enough to mislead you thinking that everything is 
all right with this residue. However, if you compute CC per atom it will 
immediately red-flag these few atoms that are not in density.

Computing CC per atom actually makes only sense if density shows you 
atoms, that is the data resolution is high enough. For example, you 
would not want to compute CC per atom at resolutions like 3A or so.

> Is there a way to tell the program to write CC one way or the other?

phenix.real_space_correlation allows more fine tuning. In particular it 
allows you to specify whether to output CC per atom or per residue.

Pavel


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://phenix-online.org/pipermail/phenixbb/attachments/20120821/d78f1bf4/attachment.htm>