I have tried both way and end up with different problem.
1) By editing the non-loop regions in loop templates. The resulted model contains a very strange loop region (the orientation of loop, which bends very high and leads to RMSD >5 for loop region backbone alone). So I tried to fix this problem by increasing the flanking conserved regions of the loops.But still I am getting the same problem.
2) By listing the sequence in aligment file as per usual Here I couldnot make the alingment for all four templates. Since the sequence length for one template is ~110 amino acids.
If I use the 2nd way by using the templates one by one keeping the each resulted model as myinitial model for modelling the next loop, Will it be a correct model ?
Please give me your valuable suggestions...
Selva
Modeller Caretaker modeller-care@salilab.org wrote: selva@pnu.edu wrote: > I want to model an antibody structure with multiple templates for > specific segments from each template based on the below alignment for > two chains seperately. Since I know the domain interface residues, I > could able to superimpose their relative orientation after modelling > seperately. > > L-CHAIN TARGET__: *AAAAAAAAAA11111AAAAAAAAAA22222AAAAAAAAAA33333AAAAAAAAAA* > L-CHAIN TEMPLATE: > *AAAAAAAAAA*xxxxx*AAAAAAAAAA*xxxxx*AAAAAAAAAA*xxxxx*AAAAAAAAAA* > LOOP-L1 TEMPLATE: xxxxxxxxxx*11111*xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx > LOOP-L2 TEMPLATE: xxxxxxxxxxxxxxxxxxxxxxxxx*22222*xxxxxxxxxxxxxxxxxxxxxxxxx > LOOP-L3 TEMPLATE: xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx*33333*xxxxxxxxxx > > > H-CHAIN TARGET__: *BBBBBBBBBB44444BBBBBBBBBB55555BBBBBBBBBB66666BBBBBBBBBB* > H-CHAIN TEMPLATE: > *BBBBBBBBBB*xxxxx*BBBBBBBBBB*xxxxx*BBBBBBBBBB*xxxxx*BBBBBBBBBB* > LOOP-H1 TEMPLATE: xxxxxxxxxx*44444*xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx > LOOP-H2 TEMPLATE: xxxxxxxxxxxxxxxxxxxxxxxxx*55555*xxxxxxxxxxxxxxxxxxxxxxxxx > LOOP-H3 TEMPLATE: xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx*66666*xxxxxxxxxx > > > I wish to know how to specify the segment positions to be modeled from > each template or > > Is there anyother easy way of modelling the same?.
You can just build a model using the four templates you have above, using essentially the alignment you show (except using - for a gap rather than x, of course). If your LOOP-* templates actually have additional structure (e.g. coordinates for the non-loop regions) then you can either edit the PDB file and take that structure out, or list that sequence in the alignment file as per usual (just align it with gaps in the target sequence, so that that structure isn't used in the model).
Ben Webb, Modeller Caretaker