That reasoning is suspicious, as you can see when you use it in another domain. You might say a certain ice sculpture really isn't just a bunch of H2O molecules, because the sculpture is beautiful; and surely the molecules aren't beautiful. That would be bad reasoning. We know the sculpture just is the molecules (what else could it be?), so we simply have to get used to the idea that a bunch of molecules can be beautiful. Likewise, we might have to get used to the idea that brain tissue is happy, if the reductionist view of mental states is generally well supported by arguments. Admittedly, that sort of talk sounded odd to me too when I first encountered the idea that the mind is the brain, but I can't say it sounds terribly odd any more.

If I have a sudden hankering to eat raspberries, it strikes me that I might want not want to eat them as a means to any end. I just want to eat some raspberries. So it's not derivative in any causal sense--it's not that I want this because it's a means of getting something else, and it's not obviously a result of some other desire. But our various desires might fall into natural classes. Wanting raspberries perhaps falls into the same class as wanting chocolate, and into a different class from wanting to get together with a friend. Wanting to eat some rasperries is an "instance" (that's the word I'd use) of some general type--it's an aesthetic desire, rather than a desire for interaction. If all humans have a set of fundamental desires like this, I think the list of them is quite long, and a lot of our specific desires fall into many categories, or resist categorization.