The answer, according to Google Images, was "No!"
Then I remembered my under-used dataset, so I set to work coming up with some kind of Venn (or Euler) diagram. It turns out that these little blighters are quite difficult to construct beyond about four overlapping sets. Indeed, it is still a fairly active area of academic research.
I took my database of 289 drinks, and found the most common ingredients:
- vermouth (sweet and dry),
- gin,
- rum (dark and light),
- whisky,
- vodka,
- lemon juice,
- bitters (including Angostura),
- lime juice,
- orange juice, and
- brandy (including cognac)
- vermouth,
- gin,
- rum,
- whisky,
- vodka, and
- brandy
It's almost symmetrical with the fortuitous but curious fact that no drink in my database contains both vodka and whisky.
A bit of time spent with Inkscape, JSON, HTML5 canvases and JavaScript allowed my to come up with an interactive web page demonstrating these intersections.