[phenixbb] Problem with Autobuild

[Thomas C. Terwilliger]

Hi Claudia,

To carry on from where you are, I would recommend fixing by hand any
things that you can see are incorrect.  If you know that the loops have
been built connecting the wrong chains then I would definitely fix that.
If the sequence has been incorrectly assigned I would fix that too.  Your
hand-fixed model can then be the starting point for another round of
AutoBuild, either with "rebuilding_in_place=False" to try and add more
backbone model, or with "rebuild_in_place=True" to just try to make a
better fit of the model that you have built.

You can then use "phenix.fit_loops" to try and fill in loops that have not
yet been built.

You might also try another approach from the beginning of model-building
by starting with the MR model without trimming to poly-Ala.

You might alternatively try the new Sculptor tools in the current PHENIX
to set up your starting model.

I am guessing that you may have a relatively low-resolution map as you
have chosen to use "helices_strands_only=True".  You might want to follow
a different path and just use "rebuild_in_place=True" from the beginning,
then cut out all the places where the starting model had the wrong
sequence, and then use fit_loops to fill in these places.  The advantage
of this approach is that rebuild_in_place=True keeps your overall model,
only adjusting where the main-chain goes and rebuilding side-chains from
the Richardson rotamer library.  If AutoBuild is doing a poor job at
building your model from scratch then this might be a useful approach.

All the best,
Tom T
>>
>> Dear List,
>>
>> Sorry fo r the easy question, but I have a problem with Autobuild.
>>
>>
>>
>> I'm trying to rebuild a scFv, composed of two domains linked by a linker.
>>
>>
>>
>> I've managed to find an MR solution with a polyalanine model deprived of
>> the loops, which is my starting model for rebuilding.
>>
>>
>>
>> I've given Autobuild three input files: the mtz, the model coordinates and
>> the sequence of the scFv, asking to model only the beta strands and
>> helices.
>>
>>
>>
>> Problem is that in the final model Autobuild seems to guess incorrectly
>> the density of two beta-strands, inserting the linker there instead of the
>> expected side chains.
>>
>>
>>
>> Any suggestions on how to correct this? Shall I manually rebuild this part
>> to avoid problems? Or is it crucial to prevent errors, for example, that
>> the model and the sequence are numbered so that they are in register?
>>
>>
>>
>> Thanks in advance,
>>
>>
>>
>> Claudia
>>
>>
>>
>>
>>
>>
>> Claudia Scotti Dipartimento di Medicina Sperimentale Sezione di Patologia
>> Generale Universita' di Pavia Piazza Botta, 10 27100 Pavia Italia Tel.
>> 0039 0382 986335/8/1 Facs 0039 0382 303673
>>
>>
>>
>>
>> _________________________________________________________________
>> Hotmail: Trusted email with powerful SPAM protection.
>> https://signup.live.com/signup.aspx?id=60969_______________________________________________
>> phenixbb mailing list
>> phenixbb at phenix-online.org
>> http://phenix-online.org/mailman/listinfo/phenixbb
>>