[phenixbb] what should I do after making an iterative-build omit map

Thomas C. Terwilliger terwilliger at lanl.gov
Fri Jul 30 12:41:45 PDT 2010

Hi Jason,

I am a little puzzled about the one chain ("A"?) in
overall_best_edited.pdb.  This file should be very similar to your input
pdb file (I think overall_best.pdb in your case).

However in any case the output PDB files from iterative_build_omit have no
significance, and should be ignored. You use your input PDB file as a
comparison with the map.

So usually you run an iterative-build omit map to make sure that your
model is not biasing your map in any serious way.

I would recommend running regular autobuild (rebuild_in_place=false first
usually, then rebuild_in_place=True) to get the best model you can. If you
have some uncertainty about model bias, then make the iterative-build omit
map as you have, and use that as a reference to check whether your model
is ok. If there are aspects of your model that the iterative_build omit
map shows to be clearly incorrect, then you might want to redo them.

If you want to be extra-sure about your model and you have experimental
phases, you can run autobuild starting with your experimental phases (no
model) and specify an iterative-build omit map...in this case your final
map is totally unbiased by your model-building, but is nevertheless
greatly improved over your starting map.

You will nearly always get the best map, however, by carrying out a
regular (not omit) autobuild run.

I hope that helps!
All the best,
Tom T

>> Hello everyone,
>> I used default setting in phenix and made an iterative-build omit map (it
>> took 3 days!!!).
>> Input: a pdb file (6 identical chains in AU); an mtz file
>> output:  a resolve_composite_map.mtz; an overall_best_edited.pdb
>> questions:
>> 1) There is only one chain in overall_best_edited.pdb,
>> the resolve_composite_map.mtz clearly indicates there are 6 chains in AU.
>> 2) What's the next step for me after obtaining resolve_composite_map.mtz?
>>         a) use phenix.refine to do refinement? if this is the case, what
>> files should I use?
>>         b) use phenix.assign_sequence to  assign sequence
>> to resolve_composite_map.mtz?
>>         c) something else?
>> Thanks in advance.
>> Jason
>> U Pitt
>> _______________________________________________
>> phenixbb mailing list
>> phenixbb at phenix-online.org
>> http://phenix-online.org/mailman/listinfo/phenixbb

More information about the phenixbb mailing list