Re: [sdpd] McMaille slight update

[Armel Le Bail <alb...@cristal.org>]

>Maybe not proteins - but there could still be quite a few
>large inorganics and organics?

It is illusive to expect to realize the most difficult step of an
SDPD in automated mode (I do not consider trivial cases here).
Remember the SDPD Round Robin 2 and these proposals of
tetragonal subcells instead of the correct cubic one.

IMHO, any cell proposal obtained in automatic mode (using default
values) has to be confirmed in manual mode. Experts may go directly
to the manual mode.

Even for a protein, no need to rescale with McMaille if proper
parameters are selected in manual mode. Tetragonal Lysozyme
example below :

Try yourself the data :

Lysozyme 
tetragonal...
! Wavelength, zeropoint, Ngrid
  1.15006 0.0000 0
! Codes for symmetry
  1 1 1 0 0 0
! W, Nind
0.20  2
!Pmin, Pmax, Vmin, Vmax, Rmin, Rmax, Rmaxref
   20. 100. 100000. 300000. 0.01 0.25 0.50
! Ntests, Nruns
  -200 20
!  2-theta   Intensity
    1.168100       100.0000
    1.656041       100.0000
    1.852174       100.0000
    1.918506       100.0000
    2.089703       100.0000
    2.345176       100.0000
    2.536523       100.0000
    2.623267       100.0000
    2.916919       100.0000
    2.992132       100.0000
    3.144491       100.0000
    3.319995       100.0000
    3.419907       100.0000
    3.458655       100.0000
    3.522905       100.0000
    3.564255       100.0000
    3.660448       100.0000
    3.712916       100.0000
    3.838720       100.0000
    3.933115       100.0000
    4.098415       100.0000
    4.151839       100.0000
    4.187484       100.0000
    4.234265       100.0000

Result :
     Tetragonal cells
     Rp     Vol     Vol/V1 Ind Nsol
     0.017 239645.594  1.00  24   8
     a        b         c      alpha  beta  gamma
   79.3952  79.3952  38.0173  90.000  90.000  90.000

Well, may be the last three digits in the volume are
exaggerated, but the cell is correct ;-).

If you expect a large cell, with a volume up to
1000000A**3, then, well, a default max volume of
2000 A**3 in automated mode will not work without
rescaling the wavelength, sure. But instead of
rescaling, is it really impossible to tell to
the program what exactly are your expected max
volume and cell parameters ? At least we can
expect that crystallographers will have a
minimal knowledge (the Bragg law and a few other
definitions including the definition of the cell
parameters and volume), and that they will know
that the d(hkl) of the first peaks on the powder
pattern can suggest what will be the magnitude
of the max cell parameters and max volume.

The only problem you may have with McMaille in
manual mode is with extremely huge cells, when the
output format will be too short and will deliver
*************** instead of a number ;-). But if this
occurs, you will not be able to solve the structure,
anyway.

Armel
   


------------------------ Yahoo! Groups Sponsor ---------------------~-->
Rent DVDs from home.
Over 14,500 titles. Free Shipping
& No Late Fees. Try Netflix for FREE!
http://us.click.yahoo.com/ArdFIC/hP.FAA/ySSFAA/UIYolB/TM
---------------------------------------------------------------------~->

 

Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/