:: Home | Print ::

Research Staff - Dr. Razif Gabdoulline

  • CONTACT INFORMATION
    Phone: +49 (0)6221 - 54 51269
    Fax:     +49 (0)6221 - 54 51483
    E-mail:  gabdoulline(at)eml.org / gabdoulline(at)gmail.com



 
 
 


Razif Gabdoulline
 

WORK INFORMATION

BIOMS - Center for Modelling and Simulation in the Biosciences
MCM group at EML Research gGmbH

Modeling macromolecular motions in the cell by Brownian dynamics simulations

Macromolecular motions and interactions in the cell are essential steps in cellular life and occur on a variety of time scales. Processes like macromolecular diffusion and transport, many types of protein-protein interactions, and protein domain rearrangement occur on timescales of milliseconds and longer. These processes cannot be described by standard molecular dynamics (MD) simulation methods. Brownian dynamics (BD) simulation is one of the methods, that allows to simulate macromolecular motions on the millisecond time scale keeping atomic level accuracy in the representation of the molecules, although not taking into account internal dynamics of macromolecules. The motions of solute molecules or particles can be described as Brownian motions, when solvent molecules are not taken into account explicitly but their influence on the solute is taken into account via random forces and friction. As a result, BD simulations model macromolecular motions with low resolution (~1 A) and larger timesteps (~ 1 picosecond for proteins under normal viscosity conditions) compared t MD simulations. Larger time scale (milliseconds) processes can be modeled by BD simulations with rigid proteins as compared to molecular dynamics simulations of macromolecules (typically nanoseconds). 

The purpose of the project is to do Brownian dynamics simulations of the interaction of macromolecules using a detailed force field including hydrophobic interactions. It is planned that the simulation programs will be developed together with necessary force field parameters, relevant to the investigated cases of protein interactions.

Currently, the work is being completed on calibrating hydrophobic interactions on the basis of the simulations of electron transfer proteins. These proteins form only short living complexes, so that a full course of their interaction can be simulated at a reasonable computational cost. Fast and accurate enough method computing hydrophobic forces is implemented. These forces are proportional to the changes of solvent accessible areas of proteins in the course of their interaction. The electron transfer between donor and acceptor sites of the proteins is modeled in details as a combination of contributions from many different possible pathways along the atomic bonds and through solvent.

Electron transfer proteins interact within millisecond time scale and simulations reproduce this time scale. Simulation results correlate with available experimental data on electron transfer rates for different proteins and under different conditions. One of the interesting results is that, in all simulated cases, the diffusion contributes to the electron transfer process only partly. A significant contribution comes from the activation barriers of electron transfer events at conformations met during diffusion. The situation is very different from the case, when the proteins are bound tightly for a long time.

The work is being done on deriving interaction properties of reduced models of proteins from interaction properties of all-atom models. The reduced models are needed to do thesimulations extending to hundreds of milliseconds. It is important to deal with realistic magnitude and spatial range of interactions, while the proteins theirselves can be represented as spheres. This is especially important because our BD simulations are designed to be conforming with multi-scale simulation methodology, where different simulations methods (molecular dynamics of all atoms ­ Brownian dynamics of rigid proteins ­ diffusion of spherical models) can be used one after another depending on the stage of the process under investigation. 




Related Research


SDA (Simulation of Diffusional Association of proteins) version 4.23

ProSAT (Protein Structure Annotation Tool)

MolSurfer (a Macromolecular Interface Navigator)

PIPSA (Protein Interaction Property Similarity Analysis) version 2.0



REFERENCES

Publications (pubmed)

Google

Google.scholar


BIOGRAPHICAL INFORMATION

Bashkortostan
FMSH # 18
Faculty of Physics, Moscow State University
Institute of Mathematical Problems of Biology RAS
Northern Illinois University
European Molecular Biology Laboratory
EML Research gGmbH
University of Heidelberg



<< back

© EML Research gGmbH