[phenixbb] [ccp4bb] Autobuild in specific area
Terwilliger, Thomas Charles
terwilliger at lanl.gov
Sat Sep 13 19:03:29 PDT 2014
This is a situation that autobuild does not address very well yet. Yes, you want rebuild_in_place=False. And yes, if you do that it will remove your model and rebuild it. It will also try to build into the remainder of the map.
In practice, you already have pretty much the best that autobuild is going to produce. Delete the protein A chains that autobuild built, and put in your original protein A. Now you have your original model plus whatever autobuild was able to build.
An alternative is: select "consider_main_chain_list=True" . This will take your input model A and at each step it will check to see if parts of that model are better than any other model that it has built. This can help for keeping the input model that you put in.
All the best,
From: CCP4 bulletin board [CCP4BB at JISCMAIL.AC.UK] on behalf of Zhang Foggy [foggyccp4 at GMAIL.COM]
Sent: Saturday, September 13, 2014 10:40 AM
To: CCP4BB at JISCMAIL.AC.UK
Subject: [ccp4bb] Autobuild in specific area
I am trying to use Autobuild to build a protein-protein complex model. I obtained partial phase solution (protein-A) through MR, and can see some density of protein-B near protein-A. I would like to use Autobuild to build protein-B model into this area. I load data file, initial density map file, sequence file (both of protein A and B), and protein A model, and run Autobuild. However, The "rebuild-in-place" function confused me a lot, no matter I set it to True or False. When I set to True, it only rebuilt protein A model without adding any additional residues; while I set it to False, it deleted the input protein-A model, and rebuilt a new one... Could somebody tell me how to build the protein B model with keeping the input protein A model in Autobuild? Thanks.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the phenixbb