We usually need a bit more information to go on, but I think I have the gist of your question.
With 3D modeling software, you would typically render out an image sequence that would contain alpha information - think of it as 4 channels instead of 3, you have the red, green, blue and how transparent each pixel is. Once brought into AE, the video layer itself would have transparent sections that allow you to see the layers behind it.
With green screen work, you are telling the system that where it sees green, you would make those pixels transparent, so you can see the layer underneath.
Both of these techniques are very common in AE, and can be used in combination.
If you want some good tutorials, most AE tutorials will have some combination of Alpha channels and keying (industry term for green screen) working. If you want a specific example, look at http://www.videocopilot.net/tutorials/mid_air_emergency/ - I remember having some keying and 3-d work in that tutorial. Plus, Video CoPilot is just a great source for tutorials to expand your AE capabilities.
(If you are trying to build a 3-D model with green screen built into it - don't. There are many many better ways.
That's a very common workflow. You would bring in talent in front of a (properly lit) green screen, key them out, then have a prerendered 3-D background on a layer behind it.
I would do a web search for green screen tutorials in AE. That will be the hardest part. Getting a good green screen shot is an art.