Working Group 9: NanoBiosystems

Working Group Lead: Peter Ortoleva, ortoleva@indiana.edu

Goals and Objectives:

The objective of the NanoBio WG is to develop computational and experimental techniques for research in phenomena in nanosystems of interest in life sciences. Systems of interest include viruses, nanoparticles interacting with cells, nanocapsules for delivery of therapeutics (e.g., drugs, siRNA, genes), nanoparticles for medical imaging, large functioning macromolecular complexes, and molecular motors. Simulation techniques used include multiscale analysis to derive equations from more fundamental physicochemical models, as well as more phenomenological approaches. A key objective is to develop simulators that are free of extensive calibration by being based on an interatomic force field. The experimental objective is to quantitatively image nanobiosystems under in vivo conditions.

This working group is also linked with Working Group 5 High Performance Computing,Computational Issues and Algorithms, and Working Group 8 Theoretical Methods.


Working Group 9 Participants

 

Presentations

Monday December 1, 2008 3pm EST - Peter Ortoleva

Presentation Abstract

Presentation Slides (33MB)

Principles of Virology and Nanomedicine Derived from Laws of Molecular Physics via All-Atom Multiscale Analysis

Principles governing the behavior of bionanosystems are derived via multiscale analysis. These systems (e.g., viruses, ribosomes, and nanocapsules for therapeutic agent delivery) are supramillion atom in scale and evolve on microsecond timescales or longer. Thus, they cannot be directly simulated via molecular dynamics. At CCVT we are using a multiscale procedure for addressing this biophysical and computational grand challenge. We have implemented our methodology as a multiscale computational platform designed for fundamental research and the computer-aided design of vaccines and drugs for viral disease or for nanocapsule delivery systems.

Our procedure involves the following elements. (1) Start with an all-atom underlying description to allow the use of an interatomic force field, and thereby avoid the need for recalibration with each new application. (2) Introduce automatically generated order parameters describing nanoscale connected subsystems or order parameter field variables for membranous and aqueous disconnected subsystems wherein organization is only maintained on-the-average. (3) Identify a smallness factor associated with the slow evolution of order parameters (much slower than the 10-14 seconds of atomic collisions/vibrations); is typically a ratio of small-to-large lengths, masses, or force constants. (4) Make the ansatz that the N-atom probability density depends on the N-atom state both directly and, via the order parameters, indirectly; in the case of order parameter fields, has a functional dependence on their profiles. (5) Insert the ansatz on into the N-atom Liouville equation and arrive at an unfolded multiscale equation in which appears explicitly, and solve this differential/functional differential multiscale Liouville equation as an asymptotic expansion in . (6) Derive an exact conservation law, and obtain a Smoluchowski-type equations for the mixed dependence/functional dependence of a reduced probability density W for the order parameters. (7) Derive the Langevin equation equivalent to this equation and simulate them numerically. This has been accomplished for the order parameters and is being implemented for the order parameter fields to enable simulation of enveloped viruses, liposomes, and cell membranes via our rigorous, calibration-free approach.

Advantages of our approach include the following. (1) The result is a calibration-free model. (2) Heuristic coupling of atomistic and continuum modeling, with associated uncertainties, is avoided. (3) As the probability distribution for atomistic fluctuations is co-evolved with the order parameters, the interscale coupling between them is preserved, hence the approach is truly multiscale. (4) Unlike for projection operator methods, our approach builds-in the multiscale character of the system from the beginning, avoiding difficult to evaluate kernels and enabling an efficient multiscale computational algorithm.

See sysbio.indiana.edu under Publications for selected downloadable papers.

 

Model Repository

See Databases for Modeling for CellX and TRND databases

Resources

nanoHUB: http://www.nanohub.org/

Table sorting checkbox
Off