I have an xray structure of a dimer protein in complex with DNA. I would like to use available computational methods in order to select amino-acids that are responsible for protein dimerization. I was thinking of using MD simulations (AMBER, GROMACS, NAMD) or schrodinger package. Has anyone done something similar before?
Hello there,
I will suggest that you make use of the software known as Discovery Studio Visualizer 3.0.,. It can be requested free of charge for academic purposes. Also when you install this software, open the protein-DNA complex in the interphase of this software. It is possible to see the amino acids that are associated to the DNA structure bringing about the dimerization of the protein. To view the amino acids sequence go to "view sequence features". I think this can be of help to you.
Do you know how could I study dimer formation?
We've studies protein protein interactions via combined molecular dynamics simulations and FRET assays (https://www.researchgate.net/publication/232249153_N-Ras_Forms_Dimers_at_POPC_Membranes) but there is numerous literature on simulation of protein dimers, e.g. regarding GPCR oligomerization etc.
After Equilibration of the Simulation System the homo-dimer interface was studied via contact matrix analysis using a program called MAXIMOBY (Cheops Modelling, Udo Höweler) which is only available commercially - but time coursed aminoacid interactions can also be monitored by open source tools like the g_hbond (http://www.gromacs.org/Documentation/Gromacs_Utilities/g_hbond) module of GROMACS - I'm sure NAMD, Amber etc. have similar modules implemented.
In my experience investigating protein-protein interfaces is a challenging thing to investigate with MD due to the high complexity and can be heavily biased by your starting structure. So it would be good to have an ensemble of simulations and also an observable effect of e.g. point mutations, protonations, ....
Article N-Ras forms dimers at POPC membranes
I am also working on PPI interface (but starting from homology models, not experimental structure).
Prior to start any MD simulation I focused on what are the putative residues the should be critical in PPI, simply starting from the initial minimized structure. Therefore, this subset of hot spots will be accurately monitored along the subsequent simulation.
For my experience, in this early stage the best tools for investigating dimers are based on two methods:
1) perform an in-silico alanine scanning (eg. with Robetta and/or FoldX);
2) estimate the dG binding energy and decompose it into per-residue contribute.
For the latter method I have adopted the MM-GBSA calculation provided with the AMBERTools. Moreover I have written an in house Perl script for the Energy Decomposition Analysis
The script needs some further testing, but it works. Soon I will upload it on my public git repo ( https://github.com/dcorrada/RICEDDARIO ).
- Badges
- Science topic