Dopamine neurotransmitter and its receptors play a critical role in the cell signaling process responsible for information transfer in neurons functioning in the nervous system. Development of improved therapeutics for such disorders as Parkinson's disease and schizophrenia would be significantly enhanced with the availability of the 3D structure for the dopamine receptors and of the binding site for dopamine and other agonists and antagonists. We report here the 3D structure of the long isoform of the human D2 dopamine receptor, predicted from primary sequence using firstprinciples theoretical and computational techniques (i.e., we did not use bioinformatic or experimental 3D structural information in predicting structures). The predicted 3D structure is validated by comparison of the predicted binding site and the relative binding affinities of dopamine, three known dopamine agonists (antiparkinsonian), and seven known antagonists (antipsychotic) in the D2 receptor to experimentally determined values. These structures correctly predict the critical residues for binding dopamine and several antagonists, identified by mutation studies, and give relative binding affinities that correlate well with experiments. The predicted binding site for dopamine and agonists is located between transmembrane (TM) helices 3, 4, 5, and 6, whereas the best antagonists bind to a site involving TM helices 2, 3, 4, 6, and 7 with minimal contacts to TM helix 5. We identify characteristic differences between the binding sites of agonists and antagonists.W ith the implication of G protein-coupled receptor (GPCR) in many diseases (1, 2), the need to solve the highresolution 3D structure of this class of integral membrane proteins to enable structure-based drug design is an important problem in structural biology. Despite the importance of solving the structure of the GPCRs, the only experimental 3D structure available for a GPCR is bovine rhodopsin. This lack of structures is because the GPCRs are bound to the membrane, making it difficult to express in sufficient quantities for crystallization.To provide structural and ligand binding information on GPCRs, we have been developing first-principles computational techniques for predicting the 3D structure of GPCRs using only the amino acid sequence (MembStruk) and for predicting binding site and binding energy of various ligands to GPCRs (HierDock). Using these techniques, we have reported the structure of olfactory receptors (3, 4), bovine rhodopsin (4, 5), and other GPCRs (4). Dopamine neurotransmitter plays a critical role in cellular signaling processes responsible for information transfer in neurons functioning in the nervous system (6, 7). Dopamine receptors (DR) belong to the superfamily of GPCRs, and to date there are five reported sequences for the human DR with multiple isoforms for each. The DRs may be subdivided based on their pharmacological behavior into the D1-like and the D2-like subfamilies, and these are ideal targets for treating schizophrenia and Parkinson's disease; th...