1) Put National Lab as your affiliation with a current address at Prodigal U?
2) Put both National Lab and Prodigal U as affiliations?
3) Put Prodigal U as your affiliation?
My colleagues at National Lab say I should do what is best for me careerwise, and don't worry about it too much. I would prefer to do #3, but I think that is misleading. I am leaning towards #2, since I did a bunch of the work at Prodigal U, but I suspect #1 is probably the proper thing to do (since all of the data originated there). Opinions?
Updated to add: I should also add that it is not uncommon for Fed scientists who move on to use only their new affiliations, since many Fed labs have irritating and/or lengthy internal review and release approval requirements. This can be reduced by minimizing the Federal affiliations of the authors (and thus the number of required approvals).