Re: [ccp4bb]: ++

[Edward Berry <eaberry@lbl.gov>]

***  For details on how to be removed from this list visit the  ***
***          CCP4 home page http://www.ccp4.ac.uk         ***



ernst@UTBC08.CM.UTEXAS.EDU wrote:
> 
>  However, when I calculate the correlation between pairs I get
>  a highest value of .75 with the rest .62-.68.  I did have to
>  fudge the cell constants because mapman requires identical
>  cell params.   Are the CC a bit low because of the fudging
>  of the cell params or am I wrong to think that all of the
>  dmmulti maps should be nearly identical?  If I am wrong
>  do I STILL need to combine the maps (or phases) somehow
>  in order to get as much information into a single map as is
>  possible?
> 
>  TIA
> 
>  Steve Ernst
In my experience 60-70 is quite high correlation for the initial 
MIR phases. The reason one needs to do averaging is that the initial 
maps are quite poor, i.e. noise is large. There's no reason to believe the 
noise will be correlated between the different maps. If the correlations 
were all in the 80-90% range it would mean the maps are nearly identical,
and therefore nearly identical to the true electron density, and you could
build your model right away with any one of them!

Or in your case with same crystal form and similar derivatives, 
some of the "noise" will be correlated. I would see the initial 
correlation as a measure of how much you have to gain by averaging. 
If the maps were nearly identical but uninterpretable, it would 
mean all the derivatives are giving you the same bad phases, 
and averaging wouldn't change them much.

After several cycles of averaging to improve phases, the CC will be much 
higher, maybe 90-95%. However you get a large increase even with the wrong 
intercrystal operators, so this just shows the procedure is capable of 
finding a set of phases consistent with the operators you give it.
The real test will of course be interpretability of the maps. 
Ed


-- 
Edward A. Berry, MailStop 3-250
Lawrence Berkeley National Laboratory
1 Cyclotron Road, Berkeley, CA 94720

Phone: +1-510-486-4335
Fax: +1-510-486-6059
Jfax +1-530-323-9836 (you send fax, I receive email)
LBNL Emergency status number: 800-445-5830 - Call after "The Big One"
	to see if the lab slid off into the bay.
e-mail: EABerry@lbl.gov
http://www.lbl.gov/LBL-Programs/pbd/xl_research/scientists/BerryE.htm
http://www.lbl.gov/~berry/
http://www.lbl.gov/~berry/berrygroup.html