hβ2R-Gαs complex: prediction versus crystal structure--how valuable are predictions based on molecular modeling studies?

In 2010, we predicted two models for the hβ(2)R-Gα(s) complex by combining the technique of homology modeling with a potential energy surface scan, since a complete crystal structure of the hβ(2)R-Gα(s) complex was not available. The crystal structure of opsin co-crystallized with part of the C-terminus of Gα (3DQB) was used as a template to model the hβ(2)R, whereas the crystal structure of Gα (1AZT) was used as a template to model Gα(s). Utilizing a potential energy surface scan between hβ(2)R and Gα(s), a six-dimensional potential energy surface was obtained. Two significant minimum regions were located on this surface, and each was associated with a distinct hβ(2)R-Gα(s) complex, namely model I and model II [Straßer A, Wittmann H-J (2010) J Mol Model 16:1307-1318]. The crystal structure of the hβ(2)R-Gα(s)βγ complex has recently been published. Thus, the aim of the current study was, on the one hand, to compare our predicted structures with the true crystal structure, and on the other to discuss the question: how valuable are predictions based on molecular modeling studies?