Title: Boltzmann Transport Monte Carlo Code - BioMOCA
1Boltzmann Transport Monte Carlo Code - BioMOCA
Reza Toghraee and Umberto Ravaioli Beckman
Institute ECE Department University of Illinois
at Urban-Champaign
- NIH Nanomedicine Center for Design of Biomimetic
Nanoconductors - Network for Computational Nanotechnology (NCN)
NanoHUB
2BioMOCA code
- Adaptation of the Monte Carlo approach used for
solid-state device simulation. - Germane to Brownian Dynamics with implicit water
description, but interaction between ions and
water is resolved using a scattering process. - BioMOCA code - developed by Trudy van der
Straaten, Gulzar Kathawala, Reza Toghraee, and
Umberto Ravaioli at the University of Illinois. - Development funded by DARPA and NSF Network for
Computational Nanotechnology.
3Ion Channels
- Biological membranes are made out of lipids to
protect the valuable interior contents of the
cell - Membrane proteins are embedded in the membrane
bilayer lipid - Ion channels answer the need for transport system
- Highly specific filters
- For instance, ion channels play a key role in
heart pulsing, neuron and muscle cells, toxins,
and are related to many diseases
http//www.rsc.org/chemsoc/
4Motivations
- Engineers trying to model channels similar to
devices - In particular we are interested in bio-inspired
structures, to realize - sensors
- artificial organs
- nano batteries
- etc
5Computational Goals and Methods
- Quantum Chemistry
- Very few particles
- Molecular Dynamics
- Extremely costly
- Limited time intervals
- Ion traversal is a rare lucky event
- A large number of ion crossings must be detected
- Monte Carlo / Brownian Dynamics
- BioMOCA is based on Transport Mante Carlo
- Continuum Models
6Transport Monte Carlo Particle Simulations
- 3D particle trajectories
- P3M
- Continuum background
- Implicit water
- Scattering
- Thermalizes ions
- Finite size of the ions
7initialize ( t 0 ) grid, protein
charges calculate ?fixed
Lennard Jones 6-12 potential
rions? ?ions
solve POISSON update E
P3M
t ? t dt
short range forces
Move ions (E FLJ) scattering with
water scattering off protein/lipid update rions
LJ
8Complexity of Monte Carlo Simulations for Typical
Ion Channels
- Very large domains and very few particles
9Complexity Continued
- Very large simulation times
- Time multi-scale problem
- Ergodicity
- MOCA is based on random numbers
10High Throughput Simulations
- IV curves
- Bias
- Ionic concentrations
- Ionic species
- Different protein configurations
- Crystallographic configurations
- Mutations
- etc
11NanoHUB and Grid Application
- How we use Grid
- i.e. OSG, or TeraGrid
- NCN account (nanoHUB)
- Demo version of Rappturized BioMOCA
- Coupling nanoHUB with the Grid
- Automatic access and job lunching for non-experts
12Recent Application
Periplasm
- Simulation of the Mechanosensitive
- Channel of Small Conductance (MscS)
- Impractical with Molecular Dynamics
- In collaboration with Prof. Klaus Schultens
group, Theoretical and Computational Biophysics
(TCB) at Beckman Institute, UIUC.
Cytoplasm
13MscS Continued
Biophysical Journal 903496-3510, 2006.
14OSG Experience with BioMOCA
- Several protein configurations of MscS channel
- Truncated protein
- Hydrated lipid
- 1800 runs on OSG
- i.e.
- Truncated protein without lipid hydration
- KCl- at 100, 200, 500 milli molar
concentrations - Filter selectivity
- 100 milli molar / 100 milli volts
- K L-gtR 0 R-gtL 5
- Cl- L-gtR 75 R-gtL 1
- Total current of 79 electron charges during 72
nano seconds