Protein Physics
Structural Change
Peptide Folding
Petascale Future
Movie Time
Computer Simulation - Basic Principles
Model System
or QM/MM
Potential
Molecular
Mechanical
Quantum
Mechanical
Molecular Mechanics Potential
V
 k b  b 
2
b
0
bonds


 k    
2
0

angles
N
  K 1  cosn      K    
2
n
0
dihedrals n 1
impropers
  
  
qq 
  4 ij  ij    ij      i j 






 rij 
i, j
 rij   i , j  Drij 

12
6
Simulation exploring the energy landscape
Energy Landscape
Product
Reactant
ANDREEA GRUIA
Halorhodopsin - Chloride Pumping at Atomic Resolution
Not Enough Room
Structure 13 617 (2005).
Valve Energy
= EI = 12 kcal/mol
Barrier without valve
= EII = 13 kcal/mol.
Sum = 25 kcal/mol
= no backflow.
.
Spring-loaded throttle valve helps prevent chloride leakage
Muscle Contraction
Thin filament
Thick filament
SONJA SCHWARZL
ATP Hydrolysis by Myosin
Biochemistry 45 5830 (2006)
BJORN WINDSHUEGEL
Dynamics of Muscle Contraction.
4/8/2015
MSBIO - Universität Heidelberg
10
PNAS 102 6873 (2005)
FRANK NOE
Large-Scale Functional
Conformational Transitions
product
reactant
s range
ns range
Ras p21
GTP-Bound Form (ON)
GDP-Bound Form (OFF)
Ras p21
GTP-Bound Form (ON)
FRANK NOE
Ras p21 trajectory
FRANK NOE
6
reactant
6
1
28
13
25
product
7
3
8
11
7
5
9
18
J. Chem. Phys. 126 155102 (2007)
Sampling:
Uniform Distribution
Sampling:
Exclusion of „Bad Structures“
Sampling:
Fail-Fast Minimization
Sampling:
Increase Density of Low-Energy Points
Transition Network:
Edges between Neighbours
Temperature
Connectivity of Network of Best Paths
Ras Molecular Switch
ON - state
OFF - state
Expensive
Toys
Spallation
Neutron
Source
Cray
XT4
One Million Atoms –
Molecular Dynamics
Cray
XT4
Lars Meinhold
Zoe Cournia
Lignocellulosic Biomass Exhibits
Structural Complexity
Hemicellulose
O
O
OAc
O
AcO
O
HO
OAc
O
HO
O
O
OH
AcO
O
O
OAc
O
O
OAc
HO
O
O
O
O-Xylan
O
O
HO
HO
O
O
O
HO
OH
OH
O
OH
HO
HO
O
HO
O
H3 CO
OH
HO
OH
Lignin
OCH3
HO
OH
HO
OH
HO
O
OCH3
H 3CO
HO
O
O
HO
HO
OCH3
HO
OCH3
OCH 3
HO
O
O
HO
OCH 3
O
OCH3
HO
OH
O
OH
OH
H 3CO
HO
O
O
O
OCH 3
OCH 3
OH
HO
OH
O
HO
Cellulose
OCH 3
HO
O
O
OCH3
OCH 3
O
Spallation
Neutron
Source
Cray
XT4
Molecular Simulation
Projects
Cellulose.
Lignin.
Lignocellulosic Biomass.
Cellulase Reaction
Mechanism.
Cellulosomes (subsequent
talk).
Computer Power Improvement for
Biological Molecular Dynamics
Simulation
Strong Scaling
TFlops
100
Now
10
1
5 years ago
0.1
10
1,000
100,000
cores
Flops
ideal
• Gromacs with Reaction-Field
• 5.4 million atoms, 175 atoms/core
Softwood Lignin & Cellulose
Cellulose Ib + 26 lignins
Probe for factors that might influence recalcitrance
 cellulose accessible surface area
 radii of gyration of lignins
Large-Scale Molecular Dynamics Simulation (1-3M atoms)
using 2008 DOE INCITE award on ORNL Cray XT4.
Cellulose:
Benjamin Lindner
Preliminary Findings
Crystalline
7 ns
Crystalline/Amorphous
1 ns
Collaborators
UT/ORNL Center for Molecular Biophysics
•
•
•
•
•
Maramuthu Krishnan (CMB, ORNL)
Loukas Petridis (CMB, ORNL)
Jiancong Xu (CMB, ORNL)
Roland Schulz (CMB, ORNL)
Benjamin Lindner (CMB, ORNL)
External
•
•
•
•
•
•
•
•
•
•
•
•
Nicoleta Bondar (U. Cal Irvine)
Lars Meinhold, Ahmed Zewail (Caltech)
Kei Moritsugu (RIKEN)
Akio Kitao (U. Tokyo)
Stefan Fischer, Isabella Daidone (U. Heidelberg)
Torsten Becker (U. Bayreuth)
Frank Noe (Free University of Berlin).
Vandana Kurkal-Siebert (BASF, Ludwigshafen).
Franci Merzel (U. Ljubljana)
John Finney (U. London)
Roy Daniel (U. Waikato)
Andrea Amadei, Alfredo Di Nola (U. Rome “La Sapienza”)