Welcome to ROSIE Rosetta Online Server that Includes Everyone

Overview: This application identifies lipid accessible residues from a protein structure and outputs a single PDB file with an adjusted B-factor column that can be used for visualization. The application does not change any protein coordinates, nor does it refine the protein structure. The application runs under 1 min for most of the proteins.

Inputs:
Input PDB
The application requires the protein to be transformed into membrane coordinates. This can be done by downloading the PDB from the PDBTM or the OPM database or by providing your own, correctly embedded PDB file. Alternatively, if you provide the PDB code, the structure will be downloaded from PDBTM.

Slice width
The application computes a two dimensional convex and concave hull from protein coordinates projected onto the xy plane. However, only coordinates within a horizontal slice with slice width are projected, from which the hulls are computed. The default slice width is 10 Å, meaning that hulls are computed for the inner and outer membrane leaflets and the space in between. Ideally, the membrane thickness of 30 Å should be an integer multiple of the slice width.

Distance cutoff
The distance cutoff defines the ‘resolution’ or ruggedness of the concave hull. Only distances larger than the cutoff are ‘cut in’ further. A large distance cutoff leads to a coarser concave hull, whereas a small distance cutoff will lead to a rugged hull. The default value is 10 Å which is approximately the distance of two Cα atoms on the same face of two neighboring helices.

Shell radius
When extending the information of the 2D concave hull back into three dimensions, we define a thicker shell than just the points (i.e. atoms) that are on the concave hull. All residues with Cα atoms within an xy distance smaller or equal than the shell radius will be part of that shell. The default shell radius is 6 Å which is about the radius of an α-helix.

What the application does on the back-end
The protein is first cleaned from hetero-atoms and then renumbered consecutively to start from 1 to bring it to Rosetta pose numbering. Then, the mp_span_from_pdb[4] application is used to create a Rosetta span file that contains the residues that are located in trans-membrane spans. Both the transformed PDB file and the newly created span file are then used to classify lipid accessibility in the membrane as described in detail in reference [2]. A single PDB file is created with an adjusted B-factor column that can then be visualized in PyMOL, together with the membrane planes.

Outputs
Span file
The span file is computed with the mp_span_from_pdb application in RosettaMP. It is used as an input file to the mp_lipid_acc application and might be useful for the user.

Model file
The PDB file has identical coordinates to the input PDB file and has adjusted B-factors. A B-factor of 50 means that the residue is lipid accessible, while a B-factor of 0 means that the residue is lipid inaccessible. The PDB file can be visualized in PyMOL to show lipid accessible residues (in orange in the image), lipid inaccessible residues (in blue), and the membrane planes. If you want to download the PDB file and manually visualize the lipid accessibility, download the membrane_planes.py, visualize_membrane.pml, and color_b-factor.pml scripts from the directory that contains the benchmark dataset (http://tinyurl.com/mp-lipid-acc-dataset). Then, open the PDB file in PyMOL, and run visualize_membrane.pml inside PyMOL to visualize the membrane and color_b-factor.pml inside PyMOL to color the protein by the B-factor column.

Command line for power users
RosettaMP’s mp_lipid_acc application requires the protein to be transformed into membrane coordinates. This can be done by downloading the PDB from the PDBTM[5] or the OPM database[6]. The PDB then needs to be cleaned and renumbered, and the Rosetta span file needs to be computed. Details of these steps can be found in the Supplement to reference[4].

The detailed command line of the mp_lipid_acc application is:

Rosetta/main/source/bin/mp_lipid_acc.macosclangrelease \
-database Rosetta/main/database \
-in:file:s 5IRX__tr.pdb \
-mp:setup:spanfiles 5IRX__tr.span \
-mp:lipid_acc:slice_width 10 \
-mp:lipid_acc:dist_cutoff 10 \
-mp:lipid_acc:shell_radius 6

where the last three options (slice_width, dist_cutoff, and shell_radius) are optional.
The application creates a PDB file with an adjusted B-factor column. The output PDB can then be visualized in PyMOL with the provided script color_b-factor.pml: open the PDB in PyMOL and then run

@color_b-factor.pml

in the viewer or command line window inside PyMOL.

Benchmark dataset
The dataset that was used to benchmark this application is available for download at http://tinyurl.com/mp-lipid-acc-dataset.

References

1. Koehler Leman, J., Lyskov, S. & Bonneau, R. Computing structure-based lipid accessibility of membrane proteins with mp_lipid_acc in RosettaMP. BMC Bioinformatics. (2016) [submitted]. bioRxiv DOI: http://dx.doi.org/10.1101/086579
   
   
2. Alford, R. F., Koehler Leman, J., Weitzner, B. D., Duran, A. M., Tilley, D. C., Elazar, A. & Gray, J. J. An Integrated Framework Advancing Membrane Protein Modeling and Design. PLoS Comput. Biol. 11, e1004398 (2015).
   
   
3. Kozma, D., Simon, I. & Tusnady, G. E. PDBTM: Protein Data Bank of transmembrane proteins after 8 years. Nucleic Acids Res. 41, D524–D529 (2013).
   
   
4. Lomize, M. A., Pogozheva, I. D., Joo, H., Mosberg, H. I. & Lomize, A. L. OPM database and PPM web server: resources for positioning of proteins in membranes. Nucleic Acids Res. 40, D370-6 (2012).
   

We welcome scientific and technical comments on our server. For support please contact us at Rosetta Forums with any comments, questions or concerns.

The mp_lipid_acc classification tool was developed by Julia Koehler Leman in the Bonneau lab. The Rosie implementation was developed by Sergey Lyskov in GrayLab@JHU.