Important: docking service is back online but is performing a low-resolution conformational search. An evolution of the service is on the way.

Description Help


G-protein-coupled receptors (GPCRs) form the largest class of membrane proteins and represents the largest gene family, accounting for 3 to 5% of the mammalian genome. They are key components of several cellular response mechanisms to environmental signals.

Class R is the only GPCR family for which structural information is currently available. The first experimental three-dimensional (3D) structure, the bovine rhodopsin, was solved in 2000. Since then, our knowledge about GPCR structures has expanded with the release of the 3D structures of the turkey β2 adrenergic receptor, squid rhodopsin , human β1 adrenergic receptor, human A2A-adenosine receptor, human dopamine D3 receptor and human CX4CR4 chemokine receptor. However, olfactory receptor structures are still missing calling for the use of homology modeling techniques.

GPCRautomodel attemps to narrow the gap between known structures and receptors of interest to the biologists. Its primary focus is the high-throughput modelization of olfactory receptor proteins and their docking against odorant molecules. The current version of GPCR automodel can potentially be employed to model any GPCR and its interaction with a molecule of interest for the user. The results are displayed in a user friendly web interface.

please cite:
G. Launay et al. Automatic Modeling of G-Protein Coupled Receptors based on Fold Recognition Techniques: Application to Mammalian Olfactory Receptors. PubMed

Receptor structure modeling Help

Jobname (optional) Email (mandatory)

Paste a protein sequence

  • Fasta format
  • Standard amino-acid one-letter code
  • Your sequence must feature seven α-helices

Choose template structure(s) Help

Human A2A adenosine receptor
Human dopamine d3 receptor
Bovine Rhodopsin (retinal form)
CX4CR4 chemokine receptor
Human B2-adrenergic receptor
Turkey B1-adrenergic receptor
Bovine Ospin (retinal free-form)Warning

Ligand binding prediction Help


The server can perform a prediction of the binding between the protein you submited and a molecule. The molecule can be picked in a database of olfactory compounds. It can alternatively be uploaded.

Enable ligand binding prediction

GPCRautomodel currently uses MODELLER for sidechains modeling.
You are required to get a licence key and paste it in the box below. The process is quick and simple.

Enter licence key