Credibility is a large hurdle you will need to overcome. No one will want to talk to you unless you have some big names backing up your research, and patents showing that you control the IP. In the science world, so much data is produced that the signal to noise ratio is high, and recognized scientists are extremely valuable. In other words, you will need to collaborate with academia, and it's probably better now than later. I suggest pitching your idea to senior faculty in different universities, especially ones focused in doing HIV stem cell research. Or perhaps working at someone's lab and setting up a collaboration with them. From what I understand, grant committees look at the name first, and then project second.
Salaries alone are nothing in the sciences, but it is expensive to set up your own lab, and could take years before you have validated that your thing actually works. For developing diagnostics, I bet there is someone trying to make a drug or therapy from CCR5. Collaborate with them and make the diagnostic part.
I know that at Johns Hopkins, anything you develop as an undergrad is 100% owned by you. Check the IP rights of other schools by talking to their tech transfer offices, and maybe consider transferring to another school.
Get someone reputable behind you, and then you guys can actually start doing something.