Dear community,
BACKGROUND:
I am homology modeling an ion channel and I'm struggling with keeping the selectivity filter of the channel from moving. I've done a lot of trouble shooting and near the end but have one last hiccup. The current step I'm on is a final relax using franklin2019 from each of the 1000 decoys outputted by rosetta homology modeling. Since the oxygens in the selectivity filter are so repulsive the filter is expanding by a significan amount. However, if I include K+ in the filter it preserves much better. So I need to find a way to automate placing K+ in very specific locations inside the selectivity filter and they need to be centered inside the 8 oxygens that make up a cube or cage for the ion. This needs to be automated because I have 1000 models to relax and it would take forever to do this manually. My solution is to create a single decoy with the appropriate ion placement then using the superimpose mover align the remaining decoys using the helices that make up the pore domain or selectivity filter, then copy and paste the coordinates from the potassium atoms. I've used this method to extract Ca2+ ions into other channels but the region I was aligning was a single continous segment.
The Problem
The residue numbering of the pore domain helices is 193-277,718-802,1243-1327,,1768-1852. If I align just one segment (193-277) or include all residues inbetween (193-1852) we have problems of favoring one side and the structure being slightly tilted or the flexible that were remodeled loops inbetween the segments pulling the structure aways from a good alignment.
Is there a way to align multiple sections of a protein that are not connected simultaniously? i.e:
9: <MOVERS>
10: <Superimpose name="align" ref_start="193,718,1243,1768" ref_end="277,802,1327,1852"
11: target_start="193,718,1243,1768" target_end="277,802,1327,1852" ref_pose="hsk2-cam-homology-from-map-refined-6cno-alignment-template-1-all-chain-A.pdb" CA_only="1"/>
12: </MOVERS>
or to make multiple movers and have teh superimposition averaged in the protocol section?
Is there another program I could use (VMD/Chimera) which I could automate with a script to do this as I know I can command line align multiple selections through those programs but there is a level of user interface and waiting for program to load I don't know how to script around.
Any directions would be super helpful thank you!
Ryan