Hi!
I am trying to build a model using the automodel class, based on two known structures where I want parts of the model to be based on structure A and parts on structure B.
My alignment file looks like this:
>P1;protein_a
structureX:protein_a:1::238:::::
aaaaaaaaaa---------aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa------------------aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa--------------------------aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa-----------------------aaaaaaaaaaaaaaaa---------*
>P1;protein_b
structureX:protein_b:1::93:::::
----------bbbbbbbbb-----------------------------------bbbbbbbbbbbbbbbbbb-------------------------------------------bbbbbbbbbbbbbbbbbbbbbbbbbb-------------------------------------------------bbbbbbbbbbbbbbbbbbbbbbb----------------bbbbbbbb*
>P1;chimera
sequence:chimera:1::238:::::
aaaaaaaaabbbbbbbbbbaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaabbbbbbbbbbbbbbbbbbbaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaabbbbbbbbbbbbbbbbbbbbbbbbbbaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaabbbbbbbbbbbbbbbbbbbbbbbaaaaaaaaaaaaaaaabbbbbbb*
In my structure files, protein_a contains residues from 1-238. The residues in protein_b are a subset of the residues in protein_a, but I want restraints to be derived from protein_b for some parts of the final model. When running this in Modeller (v 9.10) i get the error
read_te_291E> Sequence difference between alignment and pdb :
x (mismatch at alignment position 11)
which I believe comes from the fact that since protein_a contains all residues, Modeller does not skip residues in the protein_a file, and a mismatch occurs.
Do I need to remove residues from the structure in protein_a, or is there some way to tell Modeller to skip those residues when matching the sequence of residues in the structure file with the sequence of residues in the alignment file? Or alternatively tell Modeller to use restraints from only protein_b for parts of the model?
Thanks
/Per Larsosn