Multi-scale models and BIOMT

Hello, I work at Boston University working mostly with pili and helical reconstructions. Recently we have been trying to fit our subunit structures into maps generated from EM data. Because what we work with are helical in nature, we use BIOMT symmetry since it is the easiest to generate using readily available data. We wanted to see just how close our subunits get and wanted to use the MSC or Color Zone commands, but we have run into a problem with the Multi-Scale Models command. Since we are using the same BIOMT data, we assumed that Multi-Scale Models would use the same origin and orientation. This appears to be the case if we don't move the subunit structure after opening it, but after we move the subunit, the generated models don't seem to be in any relavent orientation. I have attached two pictures, both from the same view, and one showing the placement of the subunits using the 'sym' command and the other showing the surface models using Multi-Scale models. Both cases were generated off the same selected subunit. Any advice or clarification on how to use Mult-Scale Models and BIOMT information would be greatly appreciated. Be seeing you- Ken Jang -and all places are alike to me

Hi Ken, Here's what's going on. You have BIOMT matrices in your PDB file that specify the rotations and translations that operate in the coordinate system of the EM map. This is just what the Chimera "sym" command expects. The multiscale model tool though assumes that the BIOMT matrices act in the coordinate system of the PDB atoms. It knows nothing about the volume. The trouble is that you've moved the PDB model relative to the map using "Fit model in map". So the coordinate frame for the PDB atoms is no longer aligned with the coordinate frame of the EM map. In some cases this is what you want. For example if your BIOMT matrices said how to make an icosahedral virus from a monomeric subunit, you'd want it still to make the correct icosahedron if you first fit the monomer into a density map. But in your case it isn't so nice since your BIOMT matrices are defined relative to the map coordinates. Here's the fix. Fit your PDB model into the map, then use File / Save PDB..., turn on the "Save relative to mymap.mrc" option and save the new PDB. Put the BIOMT matrices in that PDB file that has already been fit into the map and multiscale will work with it. Tom
participants (2)
-
Ken Jang
-
Tom Goddard