Revisiting a Receptor-Based Pharmacophore Hypothesis for Human A(2A) Adenosine Receptor Antagonists