- Email: [email protected]
Dynamic visualization of simulated motions in a membrane ion channel
decade to perform calculations on large biomolecular systems that have now been implemented to run on a variety of computer systems, including supercomputers. Current research in this area is concerned with the characterization and validation of MD computational procedures and applications of MD to research investigations on biomolecular systems. Analysis of the tremendous amount of data produced by MD calculations on large biomolecular systems is by no means a trivial task. Nucleic acid polymers have complex structural features and even the smallest systems that one wishes to study can have several hundred atoms and several thousand degrees of freedom. Typical studies on these systems involve dynamics on the order of 100 picoseconds and require several days of dedicated processor time or many hours of supercomputer time. Detailed analysis of the data produced in these calculations is very important to fully comprehend the effect of various modeling and force-field assumptions on the results obtained. Existing analysis schemes, such as rms deviations of Cartesian coordinates and the display of several structural snapshots from the simulation, have low resolution and are limited in scope. They answer very selective questions related to overall flexibility of the molecule and fail to perform the complete analysis of the structure needed to gain a better understanding of local structural fluctuations in the molecule. A complete analysis scheme with much higher resolution will benefit future research in this area aimed at modeling more complex problems in structural biology. We have developed a novel method for a comprehensive and complete analysis of an MD calculation on a nucleic acid. The procedure, called “Dials-and-Windows,” is based on the IUPAC conformational coordinates for each nucleotide and the new, mathematically complete set of helicoidal coordinates developed by R. Lavery and H. Sklenar (called “Curves”), and is consistent with the conventions adopted at the Cambridge meeting on DNA curvature. The conformational coordinates are displayed as dials and the helicoidal coordinates as windows. The procedure also produces structural snapshots of the time evolution of the molecule in the form of a simpler “plate” structure, where only the nucleic acid bases, represented by planes, and the helical axis appear. The analysis scheme displays comprehensive structural and dynamical information on the system in a compact and rapidly accessible form. The display of results from Dials-and-Windows is based heavily on computer graphics, and the initial prototype was written to run under a VAXIVMS operating system and to generate graphical output on industry-standard Postscript printers; it was expanded subsequently to support Silicon Graphics IRIS workstations. The graphics generated by this scheme are very rich in information and greatly simplify the visualization of results and pattern recognition. The prototype Dials-and-Windows supports output files generated by three major MD programs for biomolecules: CHARM, AMBER and GROMOS. It is very important to realize that Dials-and-Windows analyzes data produced by running MD programs but does not concern itself with the setup and methodological details of MD simulations. In the past year, we have used Dials-and-Windows to analyze several computer simulation studies carried out at
44
J. Mol. Graphics,
1991, Vol. 9, March
Wesleyan. We are very pleased with the new dimension it has added to the analysis of MD results on large biomolecular systems. We have also used Dials-and-Windows as a comparison tool to study the effects of intercalating drugs on nucleic acid fragments.
Dynamic Visualization of Simulated Membrane Ion Channel
Motions in a
M. Bobak, S.-W. Chiu and E. Jakobsson Department of Physiology and Biophysics and Program In Bioengineering, University of Illinois, 156 Davenport Hall, 607 S. Matthews, Urbana, IL 61801, USA We have written a program called “MOVE” that uses the graphics library of the Silicon Graphics IRIS operating system to produce dynamic moving visualizations of molecular trajectories computed by molecular dynamic situations. The visualizations include perspective and the ability to continually orient an observed biomolecular system as it undergoes thermally driven motions critical to its function. In our laboratory MOVE has been used on a personal IRIS to produce dynamic visualizations of subpicosecond to picosecond timescale motions of the membrane ion channel peptide gramicidin-a, plus water and ions associated with the channel. In general, the visualizations provide a powerful means of developing physical intuition about the mechanisms of channel function. We use a videotape made from the visualizations to explain these phenomena at scientific meetings and university seminars. The visualizations have also provided us with a new insight into how an ion enters a channel. There appears to be a small (or no) energy barrier at the end of a channel that requires a sodium ion to lose its bulk waters of hydration before entering. ‘.’ It has been proposed that a flexible ethanolamine group at the channel mouth interacts with the ion to facilitate entry.‘.4 However, substitution of other groups for the ethanolamine appears to have no significant effect on the rate of ion entry.5.6 Visualizations of our dynamic simulations revealed an apparent explanation for this set of observations. When a sodium ion is near the channel mouth the last turn of the channel helix unwinds partly, permitting the sodium ion to interact electrostatically with carbonyl oxygens in the lumen of the channel while still retaining much of its association with bulk water. Thus the visualizations have given us a major insight into the physiological functioning of this ion channel that we would not otherwise have. REFERENCES 1 Jakobsson, E. and Chiu, S.-W. Biophys. J. 33-46 2 Chiu, S.-W. and Jakobsson, E. Biophys. J. 147-57 3 Etchebest, C. and Pullman, A. FEBS Lett. 191-5 4 Etchebest, C. and Pullman, A. FEBS Len. 261-5
1987, 52, 1987, 55, 1984, 170, 1986, 204,
5 Trudelld, Y., Daumas, P., Heitz, F., Etchebest, C. and Pullman, A. FEBS Lett. 1987, 216, 11-6 Narcessian, E.J., Anderson, O.S. and 6 Peart-Williams, Koeppe, R.E. III Biophys. J. 1988, 53, 329a
Quantum
Chemistry
for the Experimentalist
Scott D. Kahn Department of Chemistry, University of Illinois, California Street, Urbana, IL 6 1801, USA
1209 West
Theoretical developments and advances in computational hardware continue to extend the size of chemical systems that can be accurately modeled, to the point where a large part of experimental exploration may be complemented, or even replaced, with computational studies. Albeit, in the chemical sciences many of the potential benefits of “computer-aided chemistry” (CAC) may go unrealized unless tools are developed that effectively allow the experimentalist to access the computer in an intuitive and straightforward manner. Our contribution has been to develop a general graphical user interface for quantum chemistry called QMODEL. QMODEL is a comprehensive program that assists the user in building molecular structures, creating appropriate input options and submitting calculations. Upon completion of a calculation the QMODEL interface provides for the visualization of structures, orbitals, potentials and normal modes. Thus the chemist need never survey the actual numeric data unless desired; all interpretations can be made via scientific visualization techniques. At present, QMODEL has been implemented on Macintosh II and Silicon Graphics IRIS 4D graphics workstations. We discuss the design of a consistent user interface across platforms and compare desktop to deskside workstations for molecular visualization.
priate to argon, and trajectories are generated using the methods of molecular dynamics based on the velocity form of the Verlet algorithm. We have written the molecular dynamics program in QuickBASIC because the availability of graphics commands allows the user to display configuration “snapshots” as the system evolves in time. The calculation of the fractal dimension depends on one calculating the trajectory length as a function of step size, using code written in Microsoft FORTRAN. We perform all calculations on the Dell 310, a 20-MHz 80386 machine with an 80387 coprocessor. In particular, we analyze the “melting” of the cluster in terms of the fractal dimension, and we demonstrate the utility of dynamical analysis for small systems using microcomputers.
REFERENCES 1 Richardson, C.F. General Systems Yearbook 1961, 6, 139 2 Powles, J.G. and Quirke, N. Phys. Rev. Lett. 1984, 54, 1.571
Understanding the Flexibility Using Conformation Rings
of Molecular
Structures
A. R. Srinivasan, S. Back, J. L. Nauss,? S. V. Albert? and W. K. Olson Department of Chemistry, Rutgers University, New Brunswick, N.J. 08903 and t Departments of Gastroenterology and Biomedical Engineering, Walter Reed Army Institute of Research, Washington, DC 20307, USA We have developed interactive programs in FORTRAN Pascal to display conformation rings’ on Macintosh-
and and
S. G. Desjardins, Rachael M. Easton and William R. Murray Department of Chemistry, Washington and Lee University, Lexington, VA 24450, USA
270”
90”
Small clusters of atoms have received much attention in recent years because they are difficult to classify as either a solid or a liquid. In particular, it is difficult to describe such small systems (fewer than 50 particles) using thermodynamic arguments. With this in mind, we consider the cluster to be a fundamentally dynamical entity and do not use the standard approaches of statistical thermodynamics. In particular, we use the method of Richardson’,* to calculate the fractal dimension of the trajectories of single particles as a function of temperature: Particles are represented by Lennard-Jones spheres with parameters appro-
Figure I. Backbone structure
Fractal Analysis Argon Clusters
of Trajectories
in Two-Dimensional
torsions in a DNA dodecamer
J. Mol. Graphics,
1991, Vol. 9, March
crystal
45
44
J. Mol. Graphics,
1991, Vol. 9, March
Wesleyan. We are very pleased with the new dimension it has added to the analysis of MD results on large biomolecular systems. We have also used Dials-and-Windows as a comparison tool to study the effects of intercalating drugs on nucleic acid fragments.
Dynamic Visualization of Simulated Membrane Ion Channel
Motions in a
M. Bobak, S.-W. Chiu and E. Jakobsson Department of Physiology and Biophysics and Program In Bioengineering, University of Illinois, 156 Davenport Hall, 607 S. Matthews, Urbana, IL 61801, USA We have written a program called “MOVE” that uses the graphics library of the Silicon Graphics IRIS operating system to produce dynamic moving visualizations of molecular trajectories computed by molecular dynamic situations. The visualizations include perspective and the ability to continually orient an observed biomolecular system as it undergoes thermally driven motions critical to its function. In our laboratory MOVE has been used on a personal IRIS to produce dynamic visualizations of subpicosecond to picosecond timescale motions of the membrane ion channel peptide gramicidin-a, plus water and ions associated with the channel. In general, the visualizations provide a powerful means of developing physical intuition about the mechanisms of channel function. We use a videotape made from the visualizations to explain these phenomena at scientific meetings and university seminars. The visualizations have also provided us with a new insight into how an ion enters a channel. There appears to be a small (or no) energy barrier at the end of a channel that requires a sodium ion to lose its bulk waters of hydration before entering. ‘.’ It has been proposed that a flexible ethanolamine group at the channel mouth interacts with the ion to facilitate entry.‘.4 However, substitution of other groups for the ethanolamine appears to have no significant effect on the rate of ion entry.5.6 Visualizations of our dynamic simulations revealed an apparent explanation for this set of observations. When a sodium ion is near the channel mouth the last turn of the channel helix unwinds partly, permitting the sodium ion to interact electrostatically with carbonyl oxygens in the lumen of the channel while still retaining much of its association with bulk water. Thus the visualizations have given us a major insight into the physiological functioning of this ion channel that we would not otherwise have. REFERENCES 1 Jakobsson, E. and Chiu, S.-W. Biophys. J. 33-46 2 Chiu, S.-W. and Jakobsson, E. Biophys. J. 147-57 3 Etchebest, C. and Pullman, A. FEBS Lett. 191-5 4 Etchebest, C. and Pullman, A. FEBS Len. 261-5
1987, 52, 1987, 55, 1984, 170, 1986, 204,
5 Trudelld, Y., Daumas, P., Heitz, F., Etchebest, C. and Pullman, A. FEBS Lett. 1987, 216, 11-6 Narcessian, E.J., Anderson, O.S. and 6 Peart-Williams, Koeppe, R.E. III Biophys. J. 1988, 53, 329a
Quantum
Chemistry
for the Experimentalist
Scott D. Kahn Department of Chemistry, University of Illinois, California Street, Urbana, IL 6 1801, USA
1209 West
Theoretical developments and advances in computational hardware continue to extend the size of chemical systems that can be accurately modeled, to the point where a large part of experimental exploration may be complemented, or even replaced, with computational studies. Albeit, in the chemical sciences many of the potential benefits of “computer-aided chemistry” (CAC) may go unrealized unless tools are developed that effectively allow the experimentalist to access the computer in an intuitive and straightforward manner. Our contribution has been to develop a general graphical user interface for quantum chemistry called QMODEL. QMODEL is a comprehensive program that assists the user in building molecular structures, creating appropriate input options and submitting calculations. Upon completion of a calculation the QMODEL interface provides for the visualization of structures, orbitals, potentials and normal modes. Thus the chemist need never survey the actual numeric data unless desired; all interpretations can be made via scientific visualization techniques. At present, QMODEL has been implemented on Macintosh II and Silicon Graphics IRIS 4D graphics workstations. We discuss the design of a consistent user interface across platforms and compare desktop to deskside workstations for molecular visualization.
priate to argon, and trajectories are generated using the methods of molecular dynamics based on the velocity form of the Verlet algorithm. We have written the molecular dynamics program in QuickBASIC because the availability of graphics commands allows the user to display configuration “snapshots” as the system evolves in time. The calculation of the fractal dimension depends on one calculating the trajectory length as a function of step size, using code written in Microsoft FORTRAN. We perform all calculations on the Dell 310, a 20-MHz 80386 machine with an 80387 coprocessor. In particular, we analyze the “melting” of the cluster in terms of the fractal dimension, and we demonstrate the utility of dynamical analysis for small systems using microcomputers.
REFERENCES 1 Richardson, C.F. General Systems Yearbook 1961, 6, 139 2 Powles, J.G. and Quirke, N. Phys. Rev. Lett. 1984, 54, 1.571
Understanding the Flexibility Using Conformation Rings
of Molecular
Structures
A. R. Srinivasan, S. Back, J. L. Nauss,? S. V. Albert? and W. K. Olson Department of Chemistry, Rutgers University, New Brunswick, N.J. 08903 and t Departments of Gastroenterology and Biomedical Engineering, Walter Reed Army Institute of Research, Washington, DC 20307, USA We have developed interactive programs in FORTRAN Pascal to display conformation rings’ on Macintosh-
and and
S. G. Desjardins, Rachael M. Easton and William R. Murray Department of Chemistry, Washington and Lee University, Lexington, VA 24450, USA
270”
90”
Small clusters of atoms have received much attention in recent years because they are difficult to classify as either a solid or a liquid. In particular, it is difficult to describe such small systems (fewer than 50 particles) using thermodynamic arguments. With this in mind, we consider the cluster to be a fundamentally dynamical entity and do not use the standard approaches of statistical thermodynamics. In particular, we use the method of Richardson’,* to calculate the fractal dimension of the trajectories of single particles as a function of temperature: Particles are represented by Lennard-Jones spheres with parameters appro-
Figure I. Backbone structure
Fractal Analysis Argon Clusters
of Trajectories
in Two-Dimensional
torsions in a DNA dodecamer
J. Mol. Graphics,
1991, Vol. 9, March
crystal
45