Hi everyone. Can I do 3 wavelength MAD phasing with AutoSol GUI with data from multiple crystals? Is there a reference/guide that I can use? Thank you. Theresa
Hi Theresa, You can do this from a little script; I am not sure it is possible from the GUI. The best approach will depend on exactly what data you have from each crystal. I am going to suppose that you have 1 wavelength from each crystal, and the crystals are similar but have some small unit cell differences. In this case I would try to: 1. run each crystal separately as a SAD experiment (you can do this in the GUI). Maybe one of these will solve your structure and you are done. If not, hopefully at least one will find the sites that you need. If so... 2. Take the sites from #1 above ("ha_1.pdb_formatted.pdb" ) and use those as inputs to autosol with "sites_file=ha_1.pdb_formatted.pdb" or something like that. Make a little script to run 3 datasets in autosol. It will look like this, where the "group=1" etc commands will tell autosol to keep all the wavelengths separate and to run SAD phasing on each individually, then merge the resulting phase information. If you leave off the group=xx commands, then autosol will treat them all as if they are from a single crystal (you can try this too if you like, and it may work if the crystals are very similar.) autosol { seq_file = seq.dat sites_file = my_sites.pdb wavelength { group = 1 data = peak.sca lambda = .9795 atom_type = Se f_prime = -7. f_double_prime = 4 } wavelength { group = 2 data = infl.sca lambda = .9792 atom_type = Se f_prime = -7. f_double_prime = 3 } wavelength { group = 3 data = remote.sca lambda = .96 atom_type = Se f_prime = -3. f_double_prime = 2.5 } } All the best, Tom T ________________________________ From: [email protected] [[email protected]] on behalf of Theresa H [[email protected]] Sent: Friday, February 03, 2012 1:44 AM To: [email protected] Subject: [phenixbb] MAD from many crystal Hi everyone. Can I do 3 wavelength MAD phasing with AutoSol GUI with data from multiple crystals? Is there a reference/guide that I can use? Thank you. Theresa
It's not totally clear to me what the question was, but I think the
answer is "sort of". There is no simple way in the current GUI to set
up the experiment as Tom described, but what you would do is this:
1. Add the files as a single-wavelength MAD experiment.
2. When you're ready to run, click the Run button - it should pop open
a menu, and instead of choosing "Run now" you want to choose "Edit
parameters and run".
3. A text editor will open with the final AutoSol parameters, and you
need to edit each wavelength to add the group number.
I'll see if it's easy to add the group number to the GUI - I think I
had resisted this because I ran out of space for more controls, but
it's probably necessary.
On Fri, Feb 3, 2012 at 8:08 AM, Terwilliger, Thomas C
2. Take the sites from #1 above ("ha_1.pdb_formatted.pdb" ) and use those as inputs to autosol with "sites_file=ha_1.pdb_formatted.pdb" or something like that. Make a little script to run 3 datasets in autosol. It will look like this, where the "group=1" etc commands will tell autosol to keep all the wavelengths separate and to run SAD phasing on each individually, then merge the resulting phase information. If you leave off the group=xx commands, then autosol will treat them all as if they are from a single crystal (you can try this too if you like, and it may work if the crystals are very similar.)
General question for Tom (etc.): what's the advantage of doing this versus treating the three wavelengths as a single-crystal MAD experiment? -Nat
How are the phases merged in this scenario? just straight combination of the HL's? sigmaA weighted? real space average? F On Feb 3, 2012, at 9:08 AM, Terwilliger, Thomas C wrote:
"group=1" etc commands will tell autosol to keep all the wavelengths separate and to run SAD phasing on each individually, then merge the resulting phase information.
--------------------------------------------- Francis E. Reyes M.Sc. 215 UCB University of Colorado at Boulder
Hi Francis,
If you have several "groups" of data, then the phases from the several groups will be merged by addition of HL coefficients, after adjustment of phases to match the origins, and optimization of an overall scale factor on the HL coefficients.
The overall scale factor on HL coefficients is estimated using the correlation of phases between the two datasets. For example if the fom of each map were 0.5, then you would expect the correlation of phases ( "group=1" etc commands will tell autosol to keep all the wavelengths separate and to run SAD phasing on each individually, then merge the resulting phase information. ---------------------------------------------
Francis E. Reyes M.Sc.
215 UCB
University of Colorado at Boulder
_______________________________________________
phenixbb mailing list
[email protected]
http://phenix-online.org/mailman/listinfo/phenixbb
Need more information here. Is there anything different among the crystals? Or is it several sets of 3wvl MAD? What resolution you working at? How big is the ASU? The best guide here is someone local, but with more information the bb can be very helpful. F On Feb 3, 2012, at 1:44 AM, Theresa H wrote:
Hi everyone.
Can I do 3 wavelength MAD phasing with AutoSol GUI with data from multiple crystals? Is there a reference/guide that I can use?
Thank you.
Theresa _______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb
participants (4)
-
Francis E Reyes -
Nathaniel Echols -
Terwilliger, Thomas C -
Theresa H