As a follow-up:
About a week or two ago, SOLOMON worked very well at 2.4A Hg-SAD (1A) phased
map for ~1100 amino acids and 6 mercury sites. Almost 70% solvent did help
quite a bit, I am sure :)
Resolve produced comparable maps that were 'subjectively' slightly less
pretty. As a crude measure of 'goodness' I can only note that arp/warp was
able to trace the SOLOMON maps further than the RESOLVE ones - about 60%
versus 55% complete sequence.
Having said this I hasten to add that the application of NCS evened things
out, this time with about ~10% improved tracing for RESOLVE maps.
Artem
P.S. Ultimately both Resolve and SOLOMON produced high quality maps that
were readily traceable by hand and via software. Both maps made me feel
fuzzy and warm inside which was the desired outcome. In comparison, the 3.5A
maps obtained using home-source MIR did not make me feel neither fuzzy nor
warm. Non-isomorphism is a killer.
P.P.S. soaking in Hg salts produced a new space group with considerably
lower overall B (resolution went from 3.5 to 2.4A also). Mercury is indeed
the sweetest of transition metals.
-----Original Message-----
From: Clemens Vonrhein [mailto:[log in to unmask]]
Sent: Thursday, August 21, 2008 10:40 PM
To: [log in to unmask]
Cc: [log in to unmask]
Subject: Re: [ccp4bb] Running Resolve after Sharp?
Dear all,
funny, I just gave a talk partly abuot this at the IUCr computing
school in Kyoto today. Weird coincidences ..
Anyway, to explain things a bit more (although previous posts already
did most of it):
- FP/SIGFP are the 'native' amplitude and sigma
- HLA/HLB/HLC/HLD is the phase probability from SHARP as
Hendrickson-Lattmann coefficients: these are in sync with FP/SIGFP
- you can calculate an initial map in two ways:
FB/PHIB = map including (average) heavy atom contribution
Fcent/PHIcent = map of light atoms only
This is all that's needed for running DM (I don't know RESOLVE, but I
guess it is similar in the input it wants):
LABIN FP=FP SIGFP=SIGFP -
HLA=HLA HLB=HLB HLC=HLC HLD=HLD -
FDM=FDM PHIDM=PHIDM
The important points are
a) use FP/SIGFP as amplitude and sigma (even if these don't really
correspond to measured data)
b) use HLA-D for phase combination (usually you should use the
experimental phases _before_ density modification whenever you do
phase combination, if you get HLs after density modifcation these
will often be very biased)
c) use FB/PHIB or Fcent/PHIcent to calculate starting map: this is
only for the initial map, not for phase combination later during
density modification.
If you want to start from the SOLOMON result (eden_flat_XY.Zpc.mtz)
the columns are named
FPsha/SIGFPsha
HLA/HLB/HLC/HLD
FBshasol/PHIBshasol or Fcentshasol/PHIcentshasol
For density modification at normal resolution (i.e. better than about
4A or so), in my experience SOLOMON gives excellent maps ... but
doesn't handle NCS in a useable way.
For a more theoretical background check our 2003 paper (Acta D)
regarding this.
Hope that helps ...
Cheers
Clemens
On Thu, Aug 21, 2008 at 03:03:34PM -0400, [log in to unmask] wrote:
> Yes, I did that recently and it worked although I found SOLOMON maps to be
> (subjectively) better. You should probably use the 'centroid' phases from
> the eden.mtz (SHARP gurus might want to correct me :))
> Artem
>
> > Hi
> >
> > Sorry for a mixed up question about two great software.
> >
> > Has anybody tried to use Resolve after sharp / autosharp? i.e.,
> > escaping the Solomon step other than handedness determination and
> > doing the density modification (& maybe NCS / building) directly with
> > the sharp output? If so, I would appreciate any input.. I am confused
> > with the eden file..
> >
> > Cheers, Partha
> >
> >
> >
> >
> >
> > --
> > MRC National Institute for Medical Research
> > Division of Molecular Structure
> > The Ridgeway, NW7 1AA, UK
> > Email: [log in to unmask]
> > Phone: + 44 208 816 2515
> >
--
***************************************************************
* Clemens Vonrhein, Ph.D. vonrhein AT GlobalPhasing DOT com
*
* Global Phasing Ltd.
* Sheraton House, Castle Park
* Cambridge CB3 0AX, UK
*--------------------------------------------------------------
* BUSTER Development Group (http://www.globalphasing.com)
***************************************************************
|