Re: [modeller_usage] help with modelling protein-DNA complex

[Modeller Caretaker <modeller-care@ucsf.edu>]

Elliott Stollar wrote:
> I have looked at the manual and previous queries posted on this 
> reoccuring subject but am unable to get modeller to model a protein-DNA 
> complex where I treat the DNA as a rigid body.
>
> All I get back are models of the protein with no DNA present.

1. Use Modeller 9v1, which has much improved support for ligands.
2. Use a Python script to build your model - TOP has been deprecated for 
a long time.
3. See http://salilab.org/modeller/9v1/manual/node18.html for an input 
file to build a model with HETATM residues preserved as rigid bodies.

There is no need to modify your PDB file, unless it is obviously 
non-standard in some way. As long as your DNA is in HETATM records, 
Modeller should pick it up just fine.

> 2. Included in my alignment file a '.' for each BLK residue and a '/' to 
> signifiy the end of a chain
>
>  >P1;9ANT
> structureX:9ANT:FIRST :@ END : : :dros:undefined: 2.4: 0.23
> RQTYTRYQTLELEKEFHFNRYLTRRRRIEIAHALCLTERQIKIWFQNRRMKWKKENK/
> .............../
> ...............*

This part is fine...

>  >P1;1AAA
> sequence:1AAA:FIRST :@ END : : :human:undefined: 0.00: 0.00
> RTAYTNTQLLELEKEFHFNKYLCRPRRVEIAALLDLTERKVKVWFQNRRMKHKRQTQ*

Your models don't contain the DNA, because you're not listing it in the 
target sequence here. The DNA ligands and/or chain breaks also need to 
be in your target sequence. See the URL above for an example alignment.

	Ben Webb, Modeller Caretaker
--
modeller-care@ucsf.edu             http://www.salilab.org/modeller/
Modeller mail list: http://salilab.org/mailman/listinfo/modeller_usage