First you'll want to define in vitro: to some people that's a clean system with just your protein, a substrate, and water.  To others it's a petri dish full of cells.  To others it's a mouse with a xenograft, and anything short of a naturally occurring system.

Then, you'll want to explain what sort of protein interaction you're looking at.  Are two different proteins coming together or breaking apart?  You'll probably want a FRET-pairing.  Do you need to know exactly where the label is?  That will determine the chemistry you have open to you.  Does the protein size need to be maintained?  You might be able to use GFP/CFP/YFP 'tags', and express the whole protein plus fluorescent 'tag' in e. coli or directly in your cell culture system.  Do you want to see colocalization?  Then you probably want tags with very different colors that don't interact much.

And your users are probably all going to have different answers to each of those questions.  Some are going to use fluorescent proteins, others will use mant- or bodipy- labeled nucleotides, others will label their proteins with Cy.5, Alexa 488, or a host of others.  

If you just want a filter set, honestly I'd just look at what Tecan or the like install into their instruments by default, it's usually 3 or 4 filters, and they're selected to be 'close enough' to the most common readouts.  But then, people tend to choose their readouts in part to work with the instrument they have available: use a monochrometer, and they can look at whatever they want.

Thanks William - great advice!