In this page we introduce some methods, tools and related codes which are developed in our Lab.
Molecular Mechanics/Coarse-grained (MM/CG) is a hybrid model proposed in
M. Neri, et al. Phys. Rev. Lett. 95(21), 218102 (2005)
to combine both accuracy of full-atom molecular dynamics (MM) and low cost of coarse-grain (CG) simulations. In this approach, the binding site is described by an atomistic potential (the GROMOS96), whilst the rest of the protein is treated with a coarse-grain potential (the Go-like model).
In the new version of the method, applied to G-Protein Coupled Receptor/ligand complexes:
M. Leguèbe, et al. PLoS ONE 7(10), e47332 (2012)
based on Gromacs 4.5.1, two artificial potentials were introduced to avoid the water diffusing away or into the hydrophobic part of the cellular membrane. The first sends the molecules back to the system, while the second mimics the membrane action on the protein. Using this approach, the time needed to perform simulations reduces more than ten times in comparison to full-atom molecular dynamics simulations.
GOMoDo is a G-Protein Coupled Receptor (GPCR) online modeling and docking webserver:
that performs automatic homology modeling and ligand docking of GPCR receptors. It uses HHsearch package 1.5.1 for performing sequence alignment. Only GPCR templates are chosen to build 3D model of given sequence by using Modeller 9.10. The obtained 3D model can be verified also with the VADAR server, and then docked with ligands uploaded by users with both Autodock VINA or HADDOCK. Binding pockets can be predicted by the FPOCKET, and structural alignment of models needed for VINA docking is performed by LOVOALIGN
Ion Channel CG model
Coarse-Grain model of ion channels that predicts selectivity using particle-particle Coloumb interactions and the induced surface charges at the channel-protein/water boundary. Development work at the German Research School for Simulation Sciences is focused on extending the model to include structural information for the charged side chains.
The most updated code and documentation can be found on the German Research School for Simulation Sciences Redmine server at the page:
To get access this server you need to be authorized by the German Research School for Simulation Sciences IT service: send an email to German Research School for Simulation Sciences helpdesk to get the sufficient credentials.
T-Pad is a computationally efficient tool, which quantitatively analyzes protein residues’ flexibility and detect backbone conformational transitions:
R. Caliandro et al., J. Chem. Theory Comput. 8(11), 4775 (2012)
Specifically T-Pad gives a quantitative description of the intrinsic plasticity of each residue. It detects, residue-by-residue, high plasticity sites in proteins, as well as relatively rigid residues (i.e. residues with similar backbone conformations across MD and/or NMR structures), also when located on flexible protein surfaces. In addition, T-Pad reveals backbone transitions between two conformations of the Ramachandran plot from MD simulations. Finally, T-pad identifies the so-called “hinge” points for conformational changes. These allow the movement of protein domains during their binding to interactors. Hence they may play a pivotal role in molecular recognition. The tool is based on directional statistics applied to the Ramachandran angles in Molecular Dynamics (MD) trajectories and/or NMR families of structures.