Hi,

Check out this post by Tassos regarding the MATRIX weight 
http://www.dl.ac.uk/list-archive-public/ccp4bb/2005-10/msg00592.html  Good 
luck,



Hidong











Leo Chavas <[EMAIL PROTECTED]> 
Sent by: CCP4 bulletin board <[email protected]>
03/05/2007 06:27 PM
Please respond to
Leo Chavas <[EMAIL PROTECTED]>


To
[email protected]
cc

Subject
Re: [ccp4bb] weighting term in refmac






Dear Li,

   There is a value of weighting term with default 0.3 in refmac5, what is
the exact meaning of it? 

You might want to have a look at:
http://www.ccp4.ac.uk/html/refmac5/keywords/xray-principal.html#weig

What is the best value of it for different resolution
data? 

I don't know exactly, but I like to use the minimum (0.01) for "bad" data 
(that is high Rmerge, high mosaicity, low resolution), and slightly higher 
for "good" data. Anyway, personally, I'm doing several runs with different 
values of weighting, and check how it affects the geometry (respective 
rmsd values).

Does it affect the result R value much?

Well yes, if you do distort like hell your protein side chains & atoms, 
you will have everything that will fit into the electron density, 
resulting in decreasing largely your stats, but you'll have little 
biological significance (as an example: using a somehow too high value of 
weighting, a lys was disrupted in 3 parts, to fit the blobs of 
density...).

Good luck.
Leo

==========================
Chavas Leonard M.G., Ph.D.
Structural Research Center
KEK,PF. 1-1 Oho
Tuskuba, Ibaraki - Japan
---------------------------------------------
PHS: +81(0)29-864-5200 (ext: 2682)
e-mail: [EMAIL PROTECTED]
URL: http://pfweis.kek.jp/~leo
========================== 

Reply via email to