Are you asking how you can reconstitute the protein into membranes with a defined orientation, or how you can determine what the orientation of the protein is in native membranes?
For a polytopic membrane protein reconstitution, the usual method is to mix the protein, detergent, and lipids together, then remove the detergent, prompting the formation of liposomes that, hopefully, include the protein. There is no way that I know of, using this method, to influence the orientation of the protein. It may be 50% inside out and 50% rightside out, or 100% one way or the other, or something in between. The shape of the protein and the size of the liposomes can have an influence. A protein with a large extramembrane domain on one side only is more likely to end up with a preferred orientation in highly curved liposomes because of its shape. The large extramembrane domain is more likely to end up on the outside of the liposomes. A protein with small extramembrane domains is more likely to end up with an equal amount of inside-out and rightside-out orientation.
Determining the orientation of the reconstituted protein depends on some special property of the protein that can be used for the purpose. For example, if you know that the protein is N-glycosylated on its extracellular domain, then you can treat the proteoliposomes with PNGase F and see what percentage of the reconstituted protein is deglycosylated in the presence and absence of an enzyme-compatible detergent that dissolves the liposomes. Another approach is partial proteolysis experiments +/- detergent combined with Western blotting using monoclonal antibodies with epitopes known to be intracellular or extracellular. If there is a known catalytic activity associated with the protein, then the effect of the detergent on this activity can be used. Another method is the effect of detergent on chemical labeling of the protein with a probe that is not membrane permeant.
Here is a reference of my own in which P-glycoprotein, which has a large cytoplasmic domain and a much smaller extracellular domain, was reconstituted into proteoliposomes and was almost exclusively in the desirable inside-out orientation.
You can have a look at the tutorials of NAMD with Membrane proteins. If you are using GROMACS then try out KALP-15 in DPPC tutorial. I have added the links below. Hope that helps.
Hi! You can orient your protein with the OPM database (you have to upload your protein .pdb file into the PPM server you will find in that page and select the other asked options)
http://opm.phar.umich.edu/server.php
then you can build your protein+membrane system using CHARMM-GUI using the Membrane Builder page (specifying the protein/membrane system)
http://www.charmm-gui.org/?doc=input/membrane
you can at this point choose the composition, the thickness of the membran, and all you want!