Computing Multimeric Structures

Modeling structures of protein asemblies consisting of more than 2 units is currently challenging. Not many computational methods exist to do so.

Early efforts by us focused on exploiting homology information and known databases of symmetric complexes to predict structures of symmetric complexes. Chris Miles (now at UW Seattle) and Brian Olson worked towards in-silico assemby of protein complexes out of of single protein chains. A novel method (Espreso) was designed to compute symmetric homo-oligomeric protein complexes when the structure of the replicated protein monomer is known and rigid. The method exploited a structural database to propose structures of symmetric homo-oligomers that can accommodate spatial replications of a protein monomer. A fast geometric similarity procedure that avoided expensive superimposition allowed determining whether a protein monomer can fit in a deposited complex.

We are currently working towards a general framework capable of multimeric assemblies of an arbitrary number of monomers. Our ongoing efforts are building on methods currently designed to model protein dimers. Our methodology design is driven by specific applications on modeling oligomerization of GPCR-coupled receptors.

On this Project:

  • Irina Hashmi

    Daniel Veltri

    Bahar Akbal-Delibas

    Nurit Haspel (Computer Science, UMass Boston)

    Nadine Kabbani(Neuroscience, GMU)

    Amarda Shehu