Dedicated stategies for GPCR MD simulation

As the most important membrane protein,  GPCR MD simulation is a quite hot topic nowadays. However, how to assemble protein/membrane system and which forcefield to assign for the whole system are two main critical and tough task for GPCR simulations.

1. Assemble protein/membrane system

Many tools are now available for protein/membrane building including:

VMD (http://www.ks.uiuc.edu/Research/vmd/)
CHARMM-GUI (http://www.charmm-gui.org/)
Desmond System Builder (http://www.deshawresearch.com/resources_desmond.html)
g_membed (http://wwwuser.gwdg.de/~ggroenh/membed.html)
InflateGro (http://moose.bio.ucalgary.ca/index.php?page=Translate_lipdis)

VMD can support well for POPC and POPE membrane system buildingwith CHARMM27 and CHARMM36 FF. However, one haveto add additional solvent and ions into the system by tcl script from VMD tutorial.It will also need some script to merge membrane and protein. Moreover, since the lipids are not pre-equilirated,it would be necessary to equilibrate the whole system at least 20 ns before MD production.

CHARMM-GUI aims to provide more convenient way for NAMD or CHARMM simulation. It cangenerate a embeded protein/membrane system with OPM positionthrough web page interface. It even can helps to assign CHARMM CGFF for the ligand. However, there are also some obviousweeknees for it: there are so many atom clashed between lipids that we can hardly believe thelipids are pre-equilibrated which claimed by the author; the inputfile provided by CHARMM-GUI is not good enough for membrane protein simulation sincethe whole quilibration step only contains no more than 3 ns and obvious GPCR helix movement are often observed within such short time which shouldn't expected at this time scale level. So,one have to improve theprotocol by himself.

Desmond System builder tool is incorporated in Schrodinger Maestro GUI and it provides very friendly interfacefor users. It can build a OPM based position for protein/membrane systemvery easily by clicking somebottoms. It can also assign CHARMM36 FF by VIPARR tool in Desmond.

Both g_membed and InflateGro are tools within Gromacs and both of them canembed the protein into a pre-equlibrated membrane system which save lot of time for equilibration. Although g_membed is a little bit difficult than InflateGro, but the output seems to be much better than the later one.

2. Force filed

It is said that CHARMM36 FF is the best FF for lipids which is currently the only FF can reproduce lipid gel phase property.However, recent Lipid 11 FF fromlatest Amber 12 is also claimed to be as good as CHARMM36 FF, although related paper isbeing reviewed these days. Both full atom FF arequite good option for protein system simulation.

There are also other FF and methods including Gromos FF which is a united atoms FF and nowadayscoarse gain MD which use dummy sphere to represent groups and accelerate the simulation dramatically (24 coreworkstation can even achieve up to several microsecond/day). The demerit for those methods are also obvious:we gain what we paid.


3. Topology for Ligand

This is always a headache problem for many people in MD simulation.In Desmond, ligand topologycan be recognized automatically with OPLS_2005 FF. However, OPLS_2005 is only good enough for tens of ns MD simulation, it israther poor if submit to micro second MD. If wewould like to use CHARMM36 FF for protein/membrane system bound with ligand,we can generate topology from Swissparam (http://swissparam.ch/) and convert them into Desmond Viparrformat by script from Desmond 2012 ($SCHRODINGER/desmond-v31023/data/viparr/converters).This tool sometimes doesn't work well, it is said that it can be fully supported by Desmond in the nextversion of Desmond which would be released at the beginning of next year. What'sneed to mention is that CHARMM CG FF is also reachable in Desmond 2012 ($SCHRODINGER/desmond-v31023/data/viparr/ff/cgenff_base_v2b7), one can build manually with those molecular templatesif the target one is not so complicated.

CHARMM CG FF also could be obtained from (https://www.paramchem.org/). If the ligandstructure is not so complicated, it may work well. However, itsometimes may not recognize ligand bond order and so on correctly. In this way, one have to  go to CHARMM forum for helps.

Amber GAFF is definitely extremely attractiveand the primary choice for a ligand bound system. When the latest LIPID 11 FF in Amber 12 come out,Amber should be the first choice formany people especially those work with ligand.

4. Efficiency

Although the hardware of computer develops sofast nowadays that CPU update one generation almost each year, the efficiency of MD simulationsseems don't improve so much these days. For instance: no matter how many CPU we use, for a typical membrane protein simulation (132 lipids, 300aaprotein, 50,000 atoms in all) with full atom FF and typical cutoff (9-10) with PME: Gromacs can get up to20 ns/day (double precision), Amber 12 ns/day, NAMD 4ns/day. Desmond is an exception since the parralization ismuch more superior than any other MD tools, it can up to 100 ns/day with 512 CPU. It can even up to several microsecond/day in Antonwith full atom FF. It would be a wise option to use either Desmond or Gromacs, ifone would like to run hundreds of ns with full atom FF. Of course, it is also acceptable for Amber and NAMD CPU performance if the simulation only last for tens of ns.

GPU technology is developing quite fast in recent one or two years and it also bring exciting news forcomputational work especially NVIDIA CUDAaccelerations. For instance,with two GTX590, Amber 12 can get up to 20ns/day while 24 core i73.6 GHz CPU can only get 4ns/day (with intel compiler, gnu is even much slower). NAMD on the other hand, can get 5ns/day with CUDA acceleration while 24 core i7 3.6 GHz CPU can only get 0.5 ns/day. Currently, GPUcalculation is not supported in Desmond, but this feature is expected to beavailable in next version.



本文引用地址：http://blog.sciencenet.cn/blog-355217-596240.html