Comment by astazangasta

Exactly this; also, as a computational biologist I cringe at the stacking of dubious models that produced this "mapping". First is the "interactome", a terrible way to model biology, which is used to predict drug interactions, a highly dubious proposition, which is used to identify similarly-acting molecules in foods, again highly dubious - this isn't even GIGO, it's GIGIGIGO. And at the end of the day, none of this crummy methodology is going to result in any wisdom superior to "eat more kale". It's unfortunate that this is the prism certain investigators have trapped themselves in.