MD MM

Molecular Mechanics/Dynamics
simulation
Cláudio M. Soares
Protein Modelling Laboratory
ITQB NOVA
Instituto de Tecnologia Química e Biológica António Xavier
Universidade Nova de Lisboa
Oeiras
ITQB NOVA
Instituto de Tecnologia Química e Biológica António Xavier
Universidade Nova de Lisboa
Av. da República,
2780-157 Oeiras
PORTUGAL
Tel: (351-1) 21446 9259 / 21446 9610
e-mail: claudio@itqb.unl.pt
http://www.itqb.unl.pt/pm URL: http://www.itqb.unl.pt/~claudio
C.M. Soares – Physical Modelling of Biomolecules – 2023

Objectives of these classes
The objectives of this and related classes are:
•To prompt your interest on understanding living systems from an atomic and molecular
perspective
•To understand that molecular motion is fundamental for the understanding of biological
processes.
•To have an overview of the most important molecular modelling methods used today in
biosciences
•To have a sufficient understanding of their physical basis
•To have an idea of real world applications
•To be able to use these methods in a supervised way (at first)
The objectives of this class are not:
•To learn the deep details of biomolecular modelling methods

Protein modelling and Structural Bioinformatics: how is it done?
Biomolecule modelling is normally based on

•Structural information
•X-ray crystallography
•NMR
•Sequence databases
•Physical models
•Computational approaches Chemistry
Physical Biology
+ and
computational methods
Biotechnology
The aim is to understand the physical world of atoms and molecules at
the microscopic level.
Why Protein Modelling ?
• Structure-Function relationships
Interpretation of experimental results (structural for
example) using physical and empirical models
• Predict (new) molecular properties
Predict structures and (altered) functions of mutant
proteins
• Study properties not accessible to experiments
Protein dynamics in the picosecond time scale

Modelling methods.
•Purely microscopic methods. Atoms are simulated in detail
Molecular potential energy functions or force fields
•Microscopic methods that use macroscopic properties
Continuum electrostatic methods
•Statistical analysis based on molecular properties
Drug design, homology based modelling
Discrete Continuum
out
in

Proteins are not rigid structures!!!
Protein function is usually mediated by dynamic aspects, many times

resulting from the dynamic behaviour of proteins
•Catalysis
•Transport
•Small molecules (TTR and tyroid hormones)
•Across membranes (Porins)
•Electron transfer proteins (cytochrome c)
•ATP synthesis
•Many others

Time scales of molecular motion
Certain experiments can give us an idea of the time scales of molecular

motion (taken from McCammon, J. A. & Harvey, S. C. (1987). Dynamics of proteins and nucleic acids. Cambridge University Press, Cambridge.)
•Bond stretching 10-14 to 10-13 s
•Protein sidechain rotation at the surface 10-11 to 10-10 s
•Hinge bending motion of polypeptide chain segments 10-11 to 10-7 s
•Alosteric transitions, folding, etc 10-5 to 10 s

The general solution for treating molecular
and atomic properties
Schrödinger’s equation
𝐇𝚿 = 𝐸𝚿 Time independent
𝜕𝚿
𝐇𝚿 = iℏ Time dependent
𝜕𝑡
ℏ2
𝐇=− 𝛁2 + 𝐕
2𝑚
ℎ
ℏ =
2𝜋
Limited application to treat biomolecular systems
•Proteins contain thousands of atoms (and they exist in solvent)
•Instead, approximations need to be used

Born-Oppenheimer Molecular Dynamics Simulation
𝐇𝚿 = 𝐸𝚿 Time independent •Forces calculated from the Quantum
Mechanical framework
ℏ2 2 •Integration of equations of motion within the
𝐇=− 𝛁 +𝐕 Classical framework (an approximation).
2𝑚
HF/3-21G
1 ps (10-12 s)

How can we describe molecular dynamics?
Simplified models have to be used
•Empirical models, within the framework of classical physics
Molecular potential energy functions
Note: these functions cannot be derived from classical physics, only the
integration of equations of motion is done within this framework
“an empirical representation of the Born-Oppenheimer approximation, according to

which the ground state of molecules is a continuous function of the atomic co-ordinates”
in Lifson, S. & Warshel, A. (1968). Consistent force field for calculations of conformations, vibrational spectra, and enthalpies of cycloalkane
and n-alkane molecules. J.Chem.Phys. 49, 5116-5129.

What is a molecular potential energy function?
These functions quantify the energy of a molecule or set of molecules in
different conformations
Electrodynamics is neglected
Force field - if we consider the forces that we can derive from it
Molecular potential energy functions depend on the co-ordinates of all atoms
in the molecular system.
V=186.2 kcal/mol V=-51.7 kcal/mol

The “real world” solution: empirical potential energy functions
Conformational functions
van Gunsteren, W. F., Billeter, S. R., Eising, A. A., Hunenberger, P. H., Kruger, P., Mark, A. E., Scott, W. R. P. & Tironi, I. G. (1996). Biomolecular
simulation: The GROMOS96 manual and user guide. BIOMOS b.v., Zurich, Groninger. ; van Gunsteren, W. F. & Berendsen, H. J. C. (1987). Groningen
molecular simulation (GROMOS) library manual. Biomos B.V, Nijenborgh 16, 9747. AG Groningen, The Netherlands.
Nb
V( r1 ,.....,rNat ) =  1 2
K ( b − b0 ) +
2 bn n n
n =1
N
 1 2
K ( −  0 ) +
n =1 2 n n n
N
 1 2
K ( −  0 ) +
n =1 2 n n n
N
 1 
K  1+ cos( mnn − n + 
n =1 2 n
N at
 C (i, j ) C (i, j ) 

qi q j
 12 12 − 6 6 +   S( rij ) +
i j  rij rij 4 0 r rij 
Sp ecial terms
Covalent bond potential
Harmonic potential
Potential energy
1 2
K b (bn − b0 ) “Experimental”
bond potential
2 n n
b0n=Equilibrium distance
Interatomic distance
Dissociation is not allowed

Sideline: How to find covalent bond potential parameters
Calculating molecular energy using QM in different conformations
Fitting this energy to a force field function
1 2
K b (bn − b0 )
2 n n
CCSD/cc-pVQZ
CO Carbon monoxide
Bond angle potential
1 2
2
K  n ( n − 0 )
n 0n=Equilibrium angle
  

Improper dihedral angle potential
1 2
Energy
K  ( n −  0 )  0n=Equilibrium angle
2 n n 
Planar centers
C
Chiral centers
B D
 
E
A C A
B
Keeping the angle between E-D in the A-B-E plane Keeping the angle between C-D and the A-B-C plane
Dihedral angle potential
1
𝐾𝜑𝑛 1 + cos( 𝑚𝑛 𝜑𝑛 − 𝛿𝑛 )
Energy
2
mn - multiplicity ; n - phase factor 0 60 360

Dihedral angle
Ethane
A D

Three maxima and three minima
C
Triple degenerated system B

Non-bonded interactions
I.The van der Waals potential
 C12 (i, j ) C6 (i, j )  Lennard-Jones
 12 − 6 
equation
 rij rij  Energy
C12 -Repulsive Equilibrium distance
•Prevent collision
C6 - Attractive
•Dispersion forces Distance
Induced dipole-Induced dipole
Non-bonded interactions
II.The electrostatic potential - +
Can be modelled with a Coulomb law
 qq 
 i j  + +
 4 0 r rij 
Point charges placed at atomic positions

- -
Exclusions applied to atoms that establish bonded
interactions

Specificity of force field parameters
Force field parameters are specific for each atom type (not element) and
for each type of interaction
•A carbon is not simply a carbon, its type depends on the chemical

environment around it
Partial charges used in carbon atoms in the GROMOS87 force field
Peptide Carbon in  carbon  carbon  carbon

bond Carboxyl in Ser in Val in His
carbon (Asp, Glu)
Charge 0.380 0.270 0.150 0.000 -0.050

Uses for molecular potential energy functions
Many uses
•Evaluate molecular energy in different conformations
•Energy minimisation of a molecule

>To get the most probable structure for the molecule
•Monte Carlo
>Metropolis Monte Carlo for instance
•Molecular dynamics simulation
•Others
Energy minimization
Starting from a conformation, obtain
conformations with lower energy
for the molecule
•Move atoms
Search conformational space Start
Energy
•Find conformations with lower
energy
Minima in the potential energy
function hipersurface
Energy minimum
“Conformational space”

Energy minimization Leu-Gly-Trp
First conformation
0
-50
Energy (kJ/mol)
-100
-150
-200
After 500 steps -250
-300
-350
-400
0 100 200 300 400 500
Energy minimisation step

Problems of energy minimization
Start
•Lots of minima
•We can end up in the closest one
Energy
Local Energy minimum
Global Energy minimum
“Conformational space”

Modelling using molecular dynamics/mechanics techniques
The MD simulation method
Molecular Dynamics (MD) simulates how the conformation of a molecule changes
with time Classical mechanics
X-ray structure 2 Molecular potential energy function
d r
F=m 2
Nb
V( r1 ,.....,rNat ) =  12K bn ( bn − b0 ) +
n
2
n =1
dt N
 12 K n ( n − 0 ) +
n
2
F = − V( r )
n =1
N
 12K n ( n
2
− 0 ) +
n
n =1
N
 12Kn 1+cos(mnn − n  +
n =1
N at
 
  C12r (12i , j ) − C6r( i6, j ) + 4i j r
qq
  S( rij ) +
i j  ij ij 0 r ij 

Dynamic molecule Special terms
Sampling conformational
states

Ex: Force from bond potential
How to calculate forces?
Newton’s equations
(a system of n particles)
𝐫1 , ⋯ , 𝐫𝑖 , ⋯ , 𝐫𝑛 (x1,y1,z1) (x1,y1,z1)
𝑥𝑖
𝐫𝑖 = 𝑦𝑖
𝑧𝑖 V = 12 K (b − b0 ) 2
2 b = ( x1 − x2 ) 2 + ( y1 − y2 ) 2 + ( z1 − z2 ) 2
d r i (t) The same
F i = mi ai = m i V V b
=
for other
dt 2 x1 b x1 coordinates
2 b ( x1 − x2 )
d r i (t) F i =
2
= x1 b
dt mi V
= K (b − b0 )
b
F = − V ( ri , .. ., r n ) V K (b − b0 )( x1 − x2 )
Fx1 = − =−
−  V(r i ,...,r n ) x1 b
Fi =
r i This can be done for all molecular interactions
Molecular dynamics simulation: practice
A molecular system is a system with lots of particles

•There are no algebraic solutions for solving its dynamics
•Numerical approximations have to be used
These are based on numerical integration of differential equations using

a small time step t, that is typically of 0.0005ps to 0.002ps (1ps=10-12 s)
•100ps correspond to 50 000 steps of numerical integration for all

particles of the system
Heavy!!

Integration of equations of motion
𝐫1 , ⋯ , 𝐫𝑖 , ⋯ , 𝐫𝑛 A system of n particles
𝑥𝑖
𝐫𝑖 = 𝑦𝑖 Positions
𝑧𝑖
𝐯1 , ⋯ , 𝐯𝑖 , ⋯ , 𝐯𝑛 Velocities F2
F1 v2
𝐚1 , ⋯ , 𝐚𝑖 , ⋯ , 𝐚𝑛 Accelerations
v1
𝑑 2 𝐫𝑖 𝐅𝑖
= 𝐚𝑖 = F4
𝑑𝑡 𝑚𝑖
F3 v4
𝑑𝐯𝑖 𝐅𝑖
= 𝐚𝑖 = v3
𝑑𝑡 𝑚𝑖
𝑑𝐫𝑖
= 𝐯𝑖
𝑑𝑡
𝐫1 , ⋯ , 𝐫𝑖 , ⋯ , 𝐫𝑛
𝑥𝑖 The Euler integrator
𝐫𝑖 = 𝑦𝑖 𝑑𝐫𝑖
𝑧𝑖 = 𝐯𝑖
𝑑𝑡
𝐯1 , ⋯ , 𝐯𝑖 , ⋯ , 𝐯𝑛 Can be approximated by
𝐚1 , ⋯ , 𝐚𝑖 , ⋯ , 𝐚𝑛 𝐫𝑖 𝑡 + ∆𝑡 − 𝐫𝑖 𝑡
≈ 𝐯𝑖 (𝑡)
∆𝑡
F2
𝑑 2 𝐫𝑖 𝐅𝑖 F1 v2
= 𝐚𝑖 = 𝐫𝑖 𝑡 + ∆𝑡 = 𝐫𝑖 𝑡 + 𝐯𝑖 (𝑡) ∙ ∆𝑡
𝑑𝑡 𝑚𝑖
v1
𝑑𝐯𝑖 𝐅𝑖 𝐅𝑖
= 𝐚𝑖 = 𝐯𝑖 𝑡 + ∆𝑡 = 𝐯𝑖 𝑡 + ∙ ∆𝑡
𝑑𝑡 𝑚𝑖 𝑚𝑖
F4
F3 v4
𝑑𝐫𝑖 • The simplest
= 𝐯𝑖 • But unstable…
𝑑𝑡 v3
𝐫1 , ⋯ , 𝐫𝑖 , ⋯ , 𝐫𝑛
𝑥𝑖 The Leap-Frog integrator
𝐫𝑖 = 𝑦𝑖 1
𝑧𝑖 𝐫𝑖 𝑡 + ∆𝑡 = 𝐫𝑖 𝑡 + 𝐯𝑖 (𝑡 + ∆𝑡) ∙ ∆𝑡
2
𝐯1 , ⋯ , 𝐯𝑖 , ⋯ , 𝐯𝑛
1 1 𝐅𝑖
𝐯𝑖 𝑡 + ∆𝑡 = 𝐯𝑖 𝑡 − ∆𝑡 + ∙ ∆𝑡
𝐚1 , ⋯ , 𝐚𝑖 , ⋯ , 𝐚𝑛 2 2 𝑚𝑖
F2
• Just a bit more complex
𝑑 2 𝐫𝑖 𝐅𝑖 • Quite stable for the time steps F1 v2
= 𝐚𝑖 =
𝑑𝑡 𝑚𝑖 used in MD
v1
𝑑𝐯𝑖 𝐅𝑖
= 𝐚𝑖 =
𝑑𝑡 𝑚𝑖 F4
F3 v4
𝑑𝐫𝑖
= 𝐯𝑖 Taken from the GROMACS 2016 manual
𝑑𝑡 v3
Including solvent
•Implicitly
Electrostatic screening, Langevin dynamics, hydrophobic terms, etc
Smith, P. E. & Pettitt, B. M. (1994). J.Phys.Chem. 98, 9700-9711; Solmajer, T. & Mehler, E. L. (1992). Int.J.Quan.Chem.
44, 291-299; Still, W. C., Tempczyk, A., Hawley, R. C. & Hendrickson, T. (1990). J.Am.Chem.Soc. 112, 6127-6129;
Wesson, L. & Eisenberg, D. (1992). Prot.Sci. 1, 227-235.
•Explicitly
Inclusion of simulated solvent
molecules
Rc
Use of periodic
R
boundary
caixa
conditions

Temperature and pressure control
A biomolecule is not an isolated system
It exchanges heat with its surroundings 3
K NVT
= NkbT
It exists in conditions of controlled temperature 2
Simulations should reflect this! 1 N
K =  mvi2
2 i =1
There are several ways to control temperature in biomolecule simulations
•Simple velocity scaling until correct temperature is reached

•Coupling the system to an “heat bath”
Ex: Berendsen, H. J. C., Postma, J. P. M., van Gunsteren, W. F., DiNola, A. & Haak, J. R. (1984). Molecular dynamics with coupling to an external bath. J. Chem.
Phys. 81, 3684-3690.
dT (t ) 1
= 2 (T0 − T ) ; =
Heat dt 
“Bath” System •Velocity scaling
t T0
vi  vi ;  = 1+ ( − 1) using 
 T
 has dimensions of time
Temperature coupling constant Similar for Pressure
A recipe for MD
•Take an X-ray or NMR determined structure

•Energy minimize
•Solvate the system (or use vacuum)
•Set initial atomic velocities at the desired temperature
•Do the integration of equations of motion

MD limitations
With the current hardware
•Protein simulations of 1000ns are normal (>1000ns for small peptides)
1-10  (1000-10000ns or above) are feasible
This limits the type of problems that can be solved using “brute force”
•Bond stretching 0.0001 to 0.0001 ns
•Protein sidechain rotation at the surface 0.01 to 0.1 ns
•Hinge bending motion of polypeptide chain segments 0.01 to 100 ns
•Alosteric transitions, folding, etc 10000 to 10000000000 ns
We are there for fast

folders
Simulating a small system –Tyrosine in water
One Tyrosine
One Tyrosine in a box of 1213 water molecules

Tyrosine in water 0-2ps

1ps = 10-12 second

1ps = 10-12 second

1ps = 10-12 second

1ps = 10-12 second

1ps = 10-12 second

1ps = 10-12 second
Peptide dynamics: Kyatorphin
Tyr-Arg

Peptide dynamics: Kyatorphin
Kyatorphin in water 0-20ps

Studying protein motion of large protein
ABN2, an endo-1,5-α-L-arabinanase from Bacillus subtilis (443 residues long)

Arabinanase is a glycosyl hydrolase that is able to cleave the glycosidic bonds of α-1,5-L-arabinan, releasing arabino-oligosaccharides and L-arabinose
McVey et al. (2014), JBIC, 19, 505-513.


McVey et al. (2014), JBIC, 19, 505-513.


McVey et al. (2014), JBIC, 19, 505-513.

ABN2 in water
0-20ps

McVey et al. (2014), JBIC, 19, 505-513.

Faster!!!
More time
ABN2 in water
0-1000ps
0-1ns
1ns=10-9 s
McVey et al. (2014), JBIC, 19, 505-513.

Even
faster!!!
More time
ABN2 in water
0-10000ps
0-10ns
1ns=10-9 s
McVey et al. (2014), JBIC, 19, 505-513.

Comparing protein motion: exposed versus buried atoms
Exposed
ABN2 in water
0-10000ps
0-10ns
1ns=10-9 s
McVey et al. (2014), JBIC, 19, 505-513.

Buried
ABN2 in water
0-10000ps
0-10ns
1ns=10-9 s
McVey et al. (2014), JBIC, 19, 505-513.


Exposed Buried
ABN2 in water
0-10000ps
0-10ns
1ns=10-9 s
McVey et al. (2014), JBIC, 19, 505-513.

Comparing protein motion: different secondary structures
100 ns
ABN2 in water
10000-110000ps
10-110ns
1ns=10-9 s
McVey et al. (2014), JBIC, 19, 505-513.

Comparing protein motion: different secondary structures
100 ns
ABN2 in water
10000-110000ps
10-110ns
1ns=10-9 s
McVey et al. (2014), JBIC, 19, 505-513.

Studying protein motions: ns time scale
Ubiquitin
Human

Studying protein motions: ns time scale
Ubiquitin in water
0-1ns
1ns=10-9s

Studying protein motions: submicrosecond time scale
Ubiquitin in water
0-100ns=0-0.1s
1s=10-6s

Enzyme stability in ethanol/water mixtures
Structures of pseudolysin (PSL) and thermolysin (TLN)
•The protease pseudolysin (PSL)
from Pseudomonas aeruginosa
has a much higher stability in
ethanol/water (25% v/v) mixtures
than thermolysin (TLN), although
they are homologous1
•The disulfide bridge between

Cys-30 and Cys-58 is important
for the stability of PSL2
•Objective: Understand
the molecular basis of
enzyme stability in
ethanol/water mixtures
using very long
(microsecond) MD
simulations
Diana Lousa
1 Ogino et al, 1999, J. Biosci. Bioeng. 87, 61-68

2 Ogino et al, 2001, Appl. Env. Microb. 67, 942-947.

Enzyme stability in ethanol/water mixtures
R.m.s.d. Analysis
In accordance with the previous experimental findings:
▪ PSL is more stable than TLN in ethanol/water media
▪ The mutant C58G of subtilisin is less stable than the
wt
The molecular reason: the contact surface of TLN

with ethanol is larger than PSL
– Enzymes may be engineered in order to

reduce this contact surface.
Lousa et al. (2012) J. Chem. Inf. Model., 52: 465-473
Behaviour of the enzymes in eth/water (25% v/v) simulations

Study conformational changes due to reduction
Cytochrome c3
DvH

Study conformational changes due to reduction
Cytochrome c3
DvH
Fully ox – Green
Fully Red - Blue
Oliveira et al. (2005), Biophys.

J., 89, 3919-30
Ana Sofia F. Oliveira

Conformational changes in bacterial ABC transporters
Example: Sav1866 from Staphylococcus aureus

•Lipid flippase-like transporter
•ABC exporter responsible for the transport of several substrates Drug
from the cytoplasm to the exterior. Resistance
Sav1866 from S.aureus inserted into a simulated DMPC membrane Oliveira et al. (2011) Proteins, 79, 1977-90

A prototype to
eukaryotic ABC
transporters like CTFR,
where mutations are
responsible for cystic
fibrosis
Oliveira et al. (2010) J Phys Chem B., 114, 5486-96. Damas et al (2011) Prot.Sci., 20, 1220-1230
Simulating membrane proteins
Sav1866
Oliveira et al. (2011), Proteins, 79, 1977-1990

Simulating complete bacterial ABC transporters
Simulated (MD) structural changes induced by ATP-hydrolysis (ATP and ADP.PI states)
Higher frequency •MC of doxorubicin in several conformations along time
gate opening with in •Minimum energy for each Z-position
the ADP+IP state
Minimum adiabatic energy profile for doxorubicin
Helix 7
Helix 11
Helix 15
2ADP.IP For all trajectories – last 50 ns 2ATP

Oliveira et al. (2011) Proteins, 79, 1977-90
Sav1866

Studying maltose translocation through the E.coli MalFGK2E importer
Abreu et al. (2021) Sci.Rep., 11:10591

Studying maltose translocation through the E.coli MalFGK2E importer
Umbrella sampling
MD simulations
Periplasmic
direction
Cytoplasmic
direction
• ATP hydrolysis lower the energetic barriers for substrate

translocation.
• The similar ADP and Apo profiles suggest ATP hydrolysis to
Pulling MD be the main driving force for nucleotide exit.
simulations • There is a decrease of the free-energy towards the NBD
direction with a concomitant increase towards MalE
direction, contributing for the irreversibility of the
transport.
Abreu et al. (2021) Sci.Rep., 11:10591
Hydrogen metabolism in sulphate reducing bacteria
H+
Electron flow
ADP+Pi ATP Proton flow
Cytoplasm
SO42- + 9H+ HS- + 4H2O

H2
Periplasm
H2 ⇌ 2H+ + 2e- Hmc
9hcc
c3 I
c3 II
Hydrogenase

Studying molecular hydrogen permeation in
[NiFe] and [NiFeSe]-hydrogenases
• Pathways of H2 towards the active site of [NiFe]-hydrogenase (H2
diffusion)
H2
H2
Vitor H. Teixeira
Teixeira et al. (2006) Biophys.J., 91, 2035-45.

[NiFe]-hydrogenase
Hydrogenase
•GROMACS 3.1.4
•GROMOS 43A1 force

field 1 [NiFe]-hydrogenase
•Atomic charges for 100 H2 molecules
metal centres 15982 water
•single point molecules
calculations 55745 atoms in total
•Gaussian 98 and
RESP fitting
•B3LYP/6-31G(d) 6 x 9 ns long MD
6-31G(2df) for Fe simulations (to have
and Ni enough statistics)
•Ni-SIa state for the
[NiFe] centre
•[FeS] centres in the
oxidised state
•Three point H2 model of

Hunter et al (Hunter et
al. 1992. J.Chem.Phys.
97: 50–59)
•Reaction field
electrostatics

[NiFe]-hydrogenase
Hydrogenase + H2
•GROMACS 3.1.4

•Gaussian 98 and
RESP fitting
[NiFe] centre
oxidised state

97: 50–59)
•Reaction field
electrostatics

[NiFe]-hydrogenase
Hydrogenase + H2 + H2O
•GROMACS 3.1.4

•Gaussian 98 and
RESP fitting
[NiFe] centre
oxidised state

97: 50–59)
•Reaction field
electrostatics

Studying molecular H2 and O2 permeation in
[NiFe]-hydrogenase
Hydrogenase + O2 + H2O
Barbosa et al. (2020), Sci.Rep., 10, 10540

H2 diffusion: Results
Amount of H2 inside Hase H2 probability density in the
along time channels and near the active site
In the end of the simulation (9 ns),

about 50% of the initial H2 is inside
H2 fills the channels between the

active site and the surface,
approaching through the Ni side
Teixeira et al, (2006), Biophys. J, 91: 2035

[NiFeSe]-hydrogenase
Comparing a [NiFe]- with a [NiFeSe]-hydrogenase
More H2 inside [NiFeSe]-hydrogenase – Higher catalytic efficiency?
Baltazar et al. (2012), JBIC, 17, pp 543-55

[NiFeSe]-hydrogenase
Comparing a [NiFe]- with a [NiFeSe]-hydrogenase
Alternative channel for H2 access to the active site in the Dmb [NiFeSe] Hase
Baltazar et al. (2012), JBIC, 17, pp 543-55

Calculation of free energies
From statistical mechanics, the free energy of a system can be derived from
the partition function as (Zwanzig, R. W. (1954). High-Temperature equation of state by a perturbation method. I. Nonpolar gases.
J.Chem.Phys. 22, 1420-1426.):
Solution: thermodynamic integration
A = −kT ln Z S0  S1
− H( p,q)
1 =0  =0
Z=
h 3N
N!  d pd qe kT
V (  ) = (1 −  )V0 + V1   0,1

V (  )
1
This is not very practical to use
A1 − A0 =  =0
 
d
V/  
G
0  1

Calculation of free energy differences
Cutinase enantioselectivity
Thermodynamic cycle
S
2-phenylpropanol R
2-phenylpropanol S
+cutinase
(transition state)
R

Enzyme catalysis in non-aqueous solvents
Enzyme catalysis - enantioselectivity
1-phenylethanol (R) in the active site of cutinase
R- and S-enantiomers of 1-phenylethanol
Nuno Micaelo
Micaelo et al. 2005

Vidinha, et. al. (2004), Biotech. Bioeng., 85, 442-449
C.M. Soares – Physical et al. 2002of Biomolecules – 2023

MicaeloModelling
Enzyme catalysis in non-aqueous solvents
Enzyme catalysis - enantioselectivity
Changing 1-phenylethanol R to S in the active site of cutinase
300 Calculated free energy difference

R->S = 19.75 kJ/mol
200
<dV/d> kJ/mol
100
-100 In water
-200
-300
0.0 0.2 0.4 0.6 0.8 1.0
 Micaelo et al. 2002
Experimentally: only the R enantiomer is converted

Calculation: the binding free energy difference between R and S is very large
Bibliography
•First MD simulations with “hard spheres”
•Alder, B. J. & Wainwright, T. E. (1957). Phase transition for a hard sphere system. J.Chem.Phys. 27, 1208-1209.
•Alder, B. J. & Wainwright, T. E. (1959). Studies in molecular dynamics. I. General Method. J.Chem.Phys. 31, 459-466.
•Alder, B. J. & Wainwright, T. E. (1960). Studies in molecular dynamics. II. Behavior of a small number of elastic spheres. J.Chem.Phys. 33, 1439-1451.
•First simulation with a continuous potential energy function

•Rahman, A. (1964). Correlations in the motion of atoms in liquid argon. Phys.Rev. 136, A405-A411.
•First MD simulation of a protein

•McCammon, J. A., Gelin, B. R. & Karplus, M. (1977). Dynamics of folded proteins. Nature 267, 585-590.
•Karplus, M. & McCammon, J. A. (1979). Protein structural fluctuations during a period of 100ps. Nature 277, 578.
•General papers and books about Molecular Mechanics / Molecular Dynamics
•Leach, A. R. (1996). Molecular modelling. Principles and applications. Addison Wesley Longman, Essex.
•Karplus, M. & Petsko, A. (1990). Molecular dynamics simulations in biology. Nature 347, 631-639.
•van Gunsteren, W. F., Weiner, P. K. & Wilkinson, A. J. (1993). Computer simulation of biomolecular systems. Theoretical and experimental
applications. Vol. 2. ESCOM, Leiden.
•van Gunsteren, W. F. & Berendsen, H. J. C. (1990). Computer simulation of molecular dynamics: Methodology, applications, and perspectives in
chemistry. Angew. Chem. Int. 29, 992-1023.
•McCammon, J. A. & Harvey, S. C. (1987). Dynamics of proteins and nucleic acids. Cambridge University Press, Cambridge.
•Brooks, C. L., Karplus, M. & Pettitt, B. M. (1988). Proteins: A theoretical perpective of dynamics, structure, and thermodinamics, 1st edition. John wiley
&Sons, New York.
•Kollman, P. (1993). Free energy calculations: applications to chemical and biochemical phenomena. Chem.Rev. 93, 2395-2417.
•Karplus M., McCammon J. A. (2002) Molecular dynamics simulations of biomolecules. Nat.Struc.Biol. 9, 646-652
•Hansson T., Oostenbrink C., van Gunsteren W. F. (2002) Molecular dynamics simulations. Curr.Opin.Struct.Biol. 12, 190-196
•van Gunsteren, W.F., D. Bakowies, R. Baron, I. Chandrasekhar, M. Christen, X. Daura, P. Gee, D.P. Geerke, A. Glattli, P.H. Hunenberger and others.
2006. Biomolecular modeling: Goals, problems, perspectives. Angew Chem Int Ed Engl 45(25):4064-92.
•Karplus, M., J. Kuriyan. 2005. Molecular dynamics and protein function. Proc Natl Acad Sci U S A 102(19):6679-85.

MD MM

Uploaded by

Copyright:

Available Formats

MD MM

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

MD MM

Uploaded by

Copyright:

Available Formats

Molecular Mechanics/Dynamics

C.M. Soares – Physical Modelling of Biomolecules – 2023

C.M. Soares – Physical Modelling of Biomolecules – 2023

Biomolecule modelling is normally based on

C.M. Soares – Physical Modelling of Biomolecules – 2023

C.M. Soares – Physical Modelling of Biomolecules – 2023

Protein function is usually mediated by dynamic aspects, many times

C.M. Soares – Physical Modelling of Biomolecules – 2023

Certain experiments can give us an idea of the time scales of molecular

C.M. Soares – Physical Modelling of Biomolecules – 2023

•Instead, approximations need to be used

C.M. Soares – Physical Modelling of Biomolecules – 2023

“an empirical representation of the Born-Oppenheimer approximation, according to

C.M. Soares – Physical Modelling of Biomolecules – 2023

V=186.2 kcal/mol V=-51.7 kcal/mol

C.M. Soares – Physical Modelling of Biomolecules – 2023

Dissociation is not allowed

C.M. Soares – Physical Modelling of Biomolecules – 2023

mn - multiplicity ; n - phase factor 0 60 360

C.M. Soares – Physical Modelling of Biomolecules – 2023

 C12 (i, j ) C6 (i, j )  Lennard-Jones

 rij rij  Energy

C12 -Repulsive Equilibrium distance

Can be modelled with a Coulomb law

Point charges placed at atomic positions

C.M. Soares – Physical Modelling of Biomolecules – 2023

•A carbon is not simply a carbon, its type depends on the chemical

Partial charges used in carbon atoms in the GROMOS87 force field

Peptide Carbon in  carbon  carbon  carbon

C.M. Soares – Physical Modelling of Biomolecules – 2023

•Evaluate molecular energy in different conformations

•Energy minimisation of a molecule

•Molecular dynamics simulation

C.M. Soares – Physical Modelling of Biomolecules – 2023

C.M. Soares – Physical Modelling of Biomolecules – 2023

•We can end up in the closest one

Global Energy minimum

C.M. Soares – Physical Modelling of Biomolecules – 2023

Dynamic molecule Special terms

C.M. Soares – Physical Modelling of Biomolecules – 2023

A molecular system is a system with lots of particles

These are based on numerical integration of differential equations using

•100ps correspond to 50 000 steps of numerical integration for all

C.M. Soares – Physical Modelling of Biomolecules – 2023

C.M. Soares – Physical Modelling of Biomolecules – 2023

•Simple velocity scaling until correct temperature is reached

•Take an X-ray or NMR determined structure

C.M. Soares – Physical Modelling of Biomolecules – 2023

We are there for fast

One Tyrosine in a box of 1213 water molecules

Tyrosine in water 0-2ps

Tyrosine in water 0-2ps

Tyrosine in water 0-2ps

Tyrosine in water 0-20ps

Tyrosine in water 0-20ps

Tyrosine in water 0-20ps

C.M. Soares – Physical Modelling of Biomolecules – 2023

Kyatorphin in water 0-20ps

C.M. Soares – Physical Modelling of Biomolecules – 2023