The "antidepressant screens" referred to here are in mice, not humans. The Nature paper is an amazing set of experiments that provide very compelling evidence that (2R,6R)-HNK is responsible for the antidepressant effects of ketamine in mouse models of depression, but Neuroskeptic is right that it's not clear that (2R,6R)-HNK is also responsible for the antidepressant effects of ketamine in humans. We know that mouse models of depression are not necessarily great; SSRIs seem to be more effective in mouse models than they are in humans.