most research grants are at the result of preliminary studies, aggregated speculation from previous publication, and well developed story. If you stayed for 3 years and you are really comfortable with what your lab did, you probably got to see a grant unfold.
i started with the first hypothesis we worked with. followed up with the observations and then the deviations/extensions. The first conclusion that allowed for the next model. The next model's results. Next hypothesis, more intricate, and just a little bit on the work entailing that. The future of this study. If you translate all that into your experience - emphasis on your project and your role - and add clinical relevance to that (if there is any) it could make for a good story. Thats what I did. I'm just saying.
Remember this is not like an IRTA application. So I would stray from saying stuff like I ran ELISAS/blots (these things are implied). I see that a lot in applications. If you are saying you are running ELISAS (which are like a 7 hour all day process), you are just saying that you were only a technician. If you want to say something along those lines, you can be like "when I assayed for the effect of Z on chemokine/x/y production we found out..." or if you are only good for western blots (the reality is that most post docs use us for technicalities), you can say that "after knocking down X with si/shRNA, or transfecting blah blah blah, I checked for the cell expression of Q and found..."
anyway... most molec bio labs do the same thing, maybe you aren't even molecular bio. If you are clinical research, which means data collection/entry, I would focus more on the phenomenology rather than start off the description by saying "I helped enter data for...".
ok that was really roundabout and may not have answered much, just PM me, i gotta go