How to model a heptamer from a monomer?
November 10, 2010 9:35 AM Subscribe
What is the best modelling software to construct hexa- and higher ordered structures of membrane proteins if you've got a monomer structure?
I have a PDB of a protein I am interested in. It is a membrane protein and the version I have of it is a monomer. In its functional form it is an ion channel and exists either as a hexamer or a heptamer. Essentially what I would like to do is to be able to generate both the hexamer and the heptamer at varying spiral angles in the membrane and minimize them to determine which is the form that I should be targetting.
I know that this might be computationally hard, but even a program that allows me to manipulate a protein around a 6 or 7 fold mirror to assemble it into a pore will help, and then I can take over from there and generate minimized structures, but so far the only thing I have been able to do is to manually arrange them and then minimize the ensemble. Ideally I would like to place one domain and then have the computer generate the other 5 or 6 based off of the symmetry that should be present. I am more familiar with symmetry in generating the equivalent positions from small molecule crystallography, but then those were much simpler and not in all likelyhood a heptamer. (Why does nature like 7's 360 goes into 6 much easier?)
I have a PDB of a protein I am interested in. It is a membrane protein and the version I have of it is a monomer. In its functional form it is an ion channel and exists either as a hexamer or a heptamer. Essentially what I would like to do is to be able to generate both the hexamer and the heptamer at varying spiral angles in the membrane and minimize them to determine which is the form that I should be targetting.
I know that this might be computationally hard, but even a program that allows me to manipulate a protein around a 6 or 7 fold mirror to assemble it into a pore will help, and then I can take over from there and generate minimized structures, but so far the only thing I have been able to do is to manually arrange them and then minimize the ensemble. Ideally I would like to place one domain and then have the computer generate the other 5 or 6 based off of the symmetry that should be present. I am more familiar with symmetry in generating the equivalent positions from small molecule crystallography, but then those were much simpler and not in all likelyhood a heptamer. (Why does nature like 7's 360 goes into 6 much easier?)
Else. If the protein was solved crystallographically there should be symmetry information within the crystal. Can you open the pdb in something like pymol or coot and display symmetry mates? In coot you might be able to open the protein in a p6 cell to give you an indication of the hexamer.
In short you need to know how the subunits interact. In the absence if this information then what you propose is very difficult.
posted by TheOtherGuy at 10:17 AM on November 10, 2010
In short you need to know how the subunits interact. In the absence if this information then what you propose is very difficult.
posted by TheOtherGuy at 10:17 AM on November 10, 2010
Response by poster: I figured that what I was asking was rather difficult. It is an NMR structure solved in house, and I think solved under conditions where it exists as the monomer. I almost just want to be able to make something that is symmetrical that spans a range of tilt angles of the pore helicies , and minimize the structures along the way. It doesn't need to be perfect as it is going to be used for virtual high throughput screening anyways, and thats more of a guide than anything.
A protein crystallographer I asked said that PDBSET in CCP4 might be able to do it, but I don't know what I would put in for symmetrical equivalent positions to get a 7mer. I think in all likelyhood the way to proceed is to just to the vHTS on the monomer structure and screen the sites that are identified that make sense chemically.
posted by koolkat at 11:36 AM on November 10, 2010
A protein crystallographer I asked said that PDBSET in CCP4 might be able to do it, but I don't know what I would put in for symmetrical equivalent positions to get a 7mer. I think in all likelyhood the way to proceed is to just to the vHTS on the monomer structure and screen the sites that are identified that make sense chemically.
posted by koolkat at 11:36 AM on November 10, 2010
PISA: a web-based tool at the EBI might also help. I have never used it myself but my boss suggested it, and according to the site can be used to 'predict probable quaternary structures'.
posted by TheOtherGuy at 2:19 AM on November 11, 2010
posted by TheOtherGuy at 2:19 AM on November 11, 2010
This thread is closed to new comments.
Have you considered posting this question to the CCP4BB ?
There can be a bit of snark there, but you usually get some helpful advice.
Else Mefi Mail me (perhaps with the PDB accession code) and I can ask some people cleverer than me.
posted by TheOtherGuy at 10:01 AM on November 10, 2010