Distrubutions of Counterions around DNA. Molecular Dynamics Simulations Results.

    Alexander Lyubartsev and Aatto Laaksonen

    Supplementary matherial to the paper: A.P.Lyubartsev, A.Laaksonen, "Molecular Dynamics Simulation of DNA in Presence of Different Counterions" Journal of Biomolecular Structure and Dynamics, v.16. p.579 (1998)


    We have carried out a series of molecular dynamics simulations of DNA (see details below) in presence of water and three types of counterions: Li, Na and Cs. The results on spatial distributions of ions, diffusion of ions near the DNA surface and on effective (solvent-mediated) ion-DNA interaction potentials will be published soon.

    This page contains supplementary material concerning mainly spatial distribution functions (SDF). The spatial distribution function is a 3-dimensional distribution of a molecule (or of an atom of a molecule) in local coordinate system of another molecule. The presentation of spatial distribution functions is tricky.  The full representation requires four-dimensional space (three space coordinates plus one for the intensity).  The standard way to proceed is to use 3D graphics and display iso-intensity surfaces.  The GOpenMol package by Leif Laaksonen provides a nice tool to do this.  You may both rotate the 3D picture and change the intensity level interactilely. The best way to view the spatial distribution function is to have GOpenMol installed as a helper application of your Web browser. See below instructions how to get and install GOpenMol. Some animated pictures of SDF (the intensity level changes with time) are also presented on this page.

    Spatial distribution functions of ions around DNA

    To build a spatial distribution function one should fix the local coordinate system. For large molecules it may be done in different ways. Picture to the right explains how we fix the local coordinate system using four phosphates of two neighbouring base pairs.

    One can also calculate average positions of atoms of the molecule in the local coordinate system and obtain in this way "averaged molecular structure". On the pictures below we displayed the averaged structures of a two base pairs fragment  and spatial distribution functions of ions around it. As an example we displayed d(AC)d(TG) fragment, while the displayed spatial distribution functions were obtained by averaging over the all 20 pairs (10 in each direction) of our sample.

    Spatial distribution of ions around two base pairs fragment:

    Counterions Gopenmol files Animated SDF
    Li baseLi.gom (1.4Mb)   LiSDF.gif  (480 K)
    Na baseNa.gom (1.2Mb)   NaSDF.gif  (475 K)
    Cs baseCs.gom (1.4Mb)   CsSDF.gif  (433 K)


    Spatial distributions around a phosphate group

      The local coordinate system is fixed in this case by phosphorus and two free oxygens: The results on water distribution around phosphates are in perfect agreement with recent X-ray study (B.Schneider, K.Patel, H.M.Berman, Biophys.J., 75, 2422 (1998))  

    Get and setup GOpenMol

    The GOpenMol package for Windows 95/98/NT as well as for a large variety of UNIX platforms (DEC, IBM, Linux, SGI) is freely available from  http://www.csc.fi./gopenmol/

    Installation is easy. Normally one need only to download the file corresponding your platform, unpack it and run Install script.
    After completing the installation setup GOpenMol  as a helper application for your Web browser. Consult your browser manual (mime-type chemical/x-gom, extension .gom), or find some hints here.

    If your have installed GOpenMol but were not successful in setting up it as a helper application, you can download the files with spatial distribution functions onto your local disk and then open them with GOpenMol program.

    Simulation details

    The molecular dynamics simulations were performed using  MDynaMix program. The simulation cell includes a periodic 10-base pairs fragment of double-strand DNA, counterions and water molecules (500 for Li ans Cs simulations and 1050 for Na).  CHARMM force field was used for DNA and flexible SPC model for water. Other simulation features: periodic boundary conditions in all directions; Ewald summation of electrostatic interactions; double time step MD algorithm (0.2/2fs), constant-temperature - constant pressure Nose-Hoover thermostat. Simulation time 2 ns for Li and Na and 1 ns for Cs.

    Alexander Lyubartsev