I'm trying to use the BridgeChains mover in Rosettascripts to build a linker between two domains as described in Kuhlman et al 2016. As a test case, I'm just building a missing 7 residue loop in the A chain of 1Z92 using the basic protocol described in the BridgeChains documentation. I've created a pdb file that includes just the A chain, renumbered to begin at residue 1, with a TER line inserted at the break; the chain break between residues 68 and 76 looks like this:
ATOM 562 CB ALA A 68 -4.007 -39.491 3.888 1.00 66.55 A C
TER
ATOM 563 N ARG A 76 -8.535 -44.152 -3.863 1.00 60.86 A N
I then run rosetta_scripts using a basic xml script to build a 7 residue loop:
<ROSETTASCRIPTS>
<SCOREFXNS>
<ScoreFunction name="centroid_scorefunction" weights="fldsgn_cen"/>
</SCOREFXNS>
<MOVERS>
<BridgeChains name="connect" chain1="1" chain2="2" motif="7LX" overlap="0" scorefxn="centroid_scorefunction" />
</MOVERS>
<PROTOCOLS>
<Add mover_name="connect"/>
</PROTOCOLS>
</ROSETTASCRIPTS>
I expect this to build a 7-residue poly-Val linker between residues 68 and 76, but for reasons I don't understand, the output structures are missing the C-terminal Ala68 of chain 1 and the N-terminal Arg76 of chain 2, with a new 7-residue linker bridging the old residues 67-77. Any help in understanding this behavior would be much appreciated.
-Kevin