Review

Goal:
• Given a protein structure,
predict its ligand bindings
Protein-Ligand
Docking Methods Applications:
• Function prediction
• Drug discovery
Thomas Funkhouser • etc.

Princeton University
CS597A, Fall 2005

1hld

Review Protein-Ligand Docking

Questions: Questions:
• Where will the ligand bind? • Where will the ligand bind?
• Which ligand will bind? • Which ligand will bind?
• How will the ligand bind? How will the ligand bind?
• When? • When?
• Why? • Why?
• etc. • etc.

Ligand

1hld 1hld

Protein-Ligand Docking Protein-Ligand Docking

Goal: Metric:
• Given a protein and a ligand, • How well do the predicted poses and conformations
determine the pose(s) and conformation(s) match measured ones (e.g., RMSD)
minimizing the total energy of the protein-ligand complex

http://www.molsoft.com/ [Jones97]

1
Protein-Ligand Docking
Challenges:
• Predicting energetics of protein-ligand binding
• Searching space of possible poses & conformations
Protein-Ligand Docking
Challenges:
• Predicting energetics of protein-ligand binding
• Searching space of possible poses & conformations
Relative position (3 degrees of freedom)
Protein-Ligand Docking
Challenges:
• Predicting energetics of protein-ligand binding
• Searching space of possible poses & conformations
§ Relative position (3 degrees of freedom)
§ Relative orientation (3 degrees of freedom)
Rotatable bonds in ligand (n degrees of freedom)
Protein-Ligand Docking
Challenges:
Predicting energetics of protein-ligand binding
• Searching space of possible poses & conformations
Protein-Ligand Docking
Challenges:
• Predicting energetics of protein-ligand binding
• Searching space of possible poses & conformations
§ Relative position (3 degrees of freedom)
Relative orientation (3 degrees of freedom)
Protein-Ligand Docking
Challenges:
• Predicting energetics of protein-ligand binding
• Searching space of possible poses & conformations
§ Relative position (3 degrees of freedom)
§ Relative orientation (3 degrees of freedom)
§ Rotatable bonds in ligand (n degrees of freedom)
Rotatable bonds in protein (m degrees of freedom)
Side chain movements Large-scale movements
2
Outline Outline
Introduction Introduction
Scoring functions Scoring functions
• Molecular mechanics • Molecular mechanics
• Empirical functions • Empirical functions
• Knowledge-based • Knowledge-based

Searching poses & conformations Searching poses & conformations

• Systematic search • Systematic search
• Molecular dynamics • Molecular dynamics
• Simulated annealing • Simulated annealing
• Genetic algorithms • Genetic algorithms
• Incremental construction • Incremental construction
• Rotamer libraries • Rotamer libraries

Results & Discussion Results & Discussion

Scoring Functions Scoring Functions

Goal: estimate binding affinity for given Goal: estimate binding affinity for given
protein, ligand, pose, and conformations protein, ligand, pose, and conformations

HIV-1 protease complexed with an hydroxyethylene isostere inhibitor [Marsden04]

Fundamental Forces of Binding Fundamental Forces of Binding

+ +

Binding Equations: Binding Equations:

Ka [R'L'] Ka [R'L']
R +L R'L' Ka = Kd-1= R +L R'L' Ka = Kd-1=
Kd [R] [L] Kd [R] [L]
Free Energy Equations:
Go = -RTln(Ka) Go = Ho - T So

3
Fundamental Forces of Binding Fundamental Forces of Binding

+ +

Binding Equations: Binding Equations:

Ka [R'L'] Ka [R'L']
R +L R'L' Ka = Kd-1= R +L R'L' Ka = Kd-1=
Kd [R] [L] Kd [R] [L]
Free Energy Equations: Free Energy Equations:
Go = -RTln(Ka) Go = Ho -T So Go = -RTln(Ka) Go = Ho - T So

Ka : 10-2 – 10-12 M
enthalpy entropy Go : -10 to -70 kcal/mol enthalpy entropy

Factors Affecting G
o
Intramolecular Forces
Intramolecular Forces Intermolecular Forces Bond lengths (Å): H – C 1.09 C – C 1.54 C – N 1.47
(covalent) (noncovalent) H – N 1.01 C = C 1.34 C – O 1.43
• Bond lengths • Electrostatics H – O 0.96 C C 1.20
• Bond angles • Dipolar interactions
• Dihedral angles • Hydrogen bonding
• Hydrophobicity
• van der Waals

H and S work against each other in many situations.

Intramolecular Forces Intramolecular Forces

Bond lengths (Å): H – C 1.09 C – C 1.54 C – N 1.47 Bond lengths (Å): H – C 1.09 C – C 1.54 C – N 1.47
H – N 1.01 C = C 1.34 C – O 1.43 H – N 1.01 C = C 1.34 C – O 1.43
H – O 0.96 C C 1.20 H – O 0.96 C C 1.20

Bond angles: Bond angles: Dihedral angles:

H
H H H H
H C H C
109.5o 109.5o H
C N C N H
H H H H C H C
H H H H H H H H
H H H H
120o ~109.5o 120o ~109.5o
H H H H
H H
H
C C C C H H
H H H H
180o 180o
H H H
H H
H
staggered H eclipsed
H C C H H C C H

4
 Intramolecular Forces Entropic Effects of Ligand Binding
Bond lengths (Å): H – C 1.09 C – C 1.54 C – N 1.47 H H
H – N 1.01 C = C 1.34 C – O 1.43 H C
H – O 0.96 C C 1.20
C H H H
Bond angles: Dihedral angles: H H H
H C
H H H
H C H C
109.5o H
C
H
N
H C H C
H
H H C H
H H
120o
H H
~109.5o
H H H H H C H H
H H
H H
H
C C H H
H H
H C H
180o
H H H H
H H H
staggered H eclipsed
H C C H
torsional strain energy (eclipsed: 1 kcal/mol)

Entropic Effects of Ligand Binding Intermolecular Forces

H H Electrostatic interactions
H C
Dipolar interactions
C H H H
H H
Hydrogen bonding
H C
Hydrophobicity
H H C H van der Waals associations
H C H H

C H
H H
Loss of degrees of freedom upon
binding can cost 60 kcal/mol

Electrostatics Electrostatics

Charge-charge Charge-charge
• pH dependence • pH dependence

O
Charge-dipole Charge-dipole -

Charge-induced Charge-induced
dipole dipole

5
Electrostatics Coulomb’s Law

Charge-charge
• pH dependence

charge-charge
+

O
Charge-dipole -

Charge-induced +
H2 +
H2 +
H3C C C CH3
dipole C C C
- H2 - H2 - H2 -
http://www.plus2physics.com/electrostatics/study_material.asp

Electrostatics Coulombic Interactions in Ligand Binding

R O
O C H2C C
Charge-charge 4-7 kcal/mol
• pH dependence CH CH2 NH3
H2N
+
arginine
Charge-dipole O
-
OH
HO
B
O
O
Charge-induced
+ H2 + H2 +
H3C C C CH3 HO O
dipole C C C
- H2 - H2 - H2 -
HO
increasing distance dependence ligand

Cation- Interactions Interactions with Metal Ions

Na+
δ+ δ+
H H

δ+ δ+
H δ− H
H + R
N
+ δ+
Hδ H H
H

http://www.owlnet.rice.edu/~chem547/lectures/lecture04.ppt#257,5,Slide 5

6
Dipolar Interactions Dipole-Dipole Interactions

- Cl +
R
+
+

O -
O - -O

+

+ +
N

0.1 – 1 kcal/mol

-Stacking Interactions:
Hydrogen Bonding End-to-Face
H
δ+ δ+
H H
2.5-3.2 Å also O-H, F-H,
even C-H
highly directional
(130o 180o) δ+ δ+
δ−
 

H H
2-10 kcal stabilization

+ δ+
Hδ H
H

http://www.owlnet.rice.edu/~chem547/lectures/lecture04.ppt#258,6,Slide 6

-Stacking Interactions:
Hydrogen Bonding in Ligand Binding
Face to Face

H H

H H

http://www.owlnet.rice.edu/~chem547/lectures/lecture04.ppt#259,7,Slide 7

7
Salt Bridges Salt Bridges in Ligand Binding

energetics similar to H-bonds (2-10 kcal/mol)

[Klebe02] Binding of napsagatran to thrombin [Klebe02]

Hydrophobicity Hydrophobicity

+ +

Binding pocket becomes “interior” upon complexation with ligand Binding pocket becomes “interior” upon complexation with ligand
Big penalty: charged or polar groups buried but unpaired

Hydrophobicity van der Waals Interactions

Dipole – induced dipole Induced dipole – induced
dipole
+
+ H2 + H2 +
O H3C C C CH3
Binding pocket becomes “interior” upon complexation with ligand - C C C
- - - -
H2 H2 H2
Big penalty: charged or polar groups buried but unpaired
+ H2 + H2 +
Energetic contribution is propotional to the size of the surface buried + H2 + H2 +
H3C C C CH3 H3C C C CH3
upon ligand binding - C -C -C - C C C
- -H -H -
• e.g. –CH3 group (25 Å2): 3 to 6 kcal/mol H2 H2 H2 H2 2 2

Distance dependent Short-range repulsion

stabilization > 1 kcal/mol

8
 van der Waals Interactions Solvation and Desolvation

AutoDock User’s Manual

AutoDock User’s Manual

Solvent Effects: Solvation and Desolvation

Hydrogen Bonding

Water
• tetrahedrally connected
network
• very small volume
Enthalpy:
• 3 – 4 H-bonds at 2 – 10
kcal/mol each

Entropy:
• flexible matrix
(disorder)

Rupture H-bonds within water matrix Reform H-bonds

Reorganize water molecules at surface Bury a hydrophobic pocket surface
http://www.bgsu.edu/departments/chem/midden/MITBCT/chem/review.html#Hbonds
Lose degrees of freedom Some water molecules released
AutoDock User’s Manual

Solvent Effects:
Solvent-Assisted Binding
Solvent-Assisted Binding

“interstitial waters”

[Klebe02]

9
Solvent Displacement Protein Conformation

Fewer degrees of freedom

New interaction surfaces
Usually driven by hydrophobicity

Scoring Functions Scoring Functions

Molecular mechanics force fields: Molecular mechanics force fields:
• CHARMM [Brooks83] • CHARMM [Brooks83]
• AMBER [Cornell95] • AMBER [Cornell95]

Empirical methods: Empirical methods:

• ChemScore [Eldridge97] • ChemScore [Eldridge97]
• GlideScore [Friesner04] • GlideScore [Friesner04]
• AutoDock [Morris98] • AutoDock [Morris98]

Knowledge-based methods Knowledge-based methods

• PMF [Muegge99] • PMF [Muegge99]
• Bleep [Mitchell99] • Bleep [Mitchell99]
• DrugScore [Gohlke00] • DrugScore [Gohlke00]

Molecular Mechanics Force Fields Molecular Mechanics Force Fields

CHARMM: AMBER:

[Brooks83] [Cornell95]

10
Scoring Functions Empirical Scoring Functions
Molecular mechanics force fields: ChemScore:
• CHARMM [Brooks83]
• AMBER [Cornell95] ∆Gbind = ∆G0 +
∆Ghbond g1 ( ∆r ) g 2 ( ∆α ) +
Empirical methods: iI

∆Gmetal f ( raM ) +
• ChemScore [Eldridge97] aM

• GlideScore [Friesner04] ∆Glipo f ( riL ) +

• AutoDock [Morris98] iL [Marsden04]
∆Grot H rot
Knowledge-based methods
• PMF [Muegge99]
• Bleep [Mitchell99]
• DrugScore [Gohlke00]
[Eldridge97]

Empirical Scoring Functions Empirical Scoring Functions

GlideScore: AutoDock 3.0:
∆Gbind = Chbond − neut−neut g ( ∆r )h( ∆α ) +
∆Gbinding = ∆GvdW + ∆Gelec + ∆Ghbond +
g ( ∆r )h( ∆α ) +
Chbond − neut−ch arg ed
∆Gdesolv + ∆Gtors
Chbond − ch arg ed −ch arg ed g ( ∆r )h( ∆α ) + ∆GvdW
C max −metal−ion f ( rlm )h( ∆α ) + 12-6 Lennard-Jones potential

Clipo − lipo f ( rlr ) + ∆Gelec

Coulombic with Solmajer-dielectric
C rotb H rotb +
∆Ghbond
C polar − phobV polar − phob 12-10 Potential with Goodford Directionality
Ccoul Ecoul + ∆Gdesolv
CvdW EvdW + Stouten Pairwise Atomic Solvation Parameters
solvationterms ∆Gtors
Number of rotatable bonds

[Friesner04] [Huey & Morris, AutoDock & ADT Tutorial, 2005]

Scoring Functions Knowledge-Based Scoring Functions

Molecular mechanics force fields: DrugScore:
• CHARMM [Brooks83]
• AMBER [Cornell95] ∆Wi ( SAS , SAS0 ) +
∆Wi , j ( r ) + (1 − γ ) ×
ki
∆W = γ
Empirical methods: ki lj ∆Wj ( SAS, SAS0 )
lj
• ChemScore [Eldridge97]
• GlideScore [Friesner04]
Protein-Ligand Solvent
• AutoDock [Morris98] Atom Accessible
Distance Surface
Knowledge-based methods Term Terms
• PMF [Muegge99]
• Bleep [Mitchell99]
• DrugScore [Gohlke00]
[Gohlke00]

11
Scoring Functions Computing Scoring Functions
Molecular mechanics force fields: Point-based calculation:
• CHARMM [Brooks83] • Sum terms computed at positions of ligand atoms
• AMBER [Cornell95] (this will be slow)

Empirical methods: p
• ChemScore [Eldridge97] How to compute Xi
r

• GlideScore [Friesner04] score efficiently?

• AutoDock [Morris98]

Knowledge-based methods
• PMF [Muegge99]
• Bleep [Mitchell99]
• DrugScore [Gohlke00]

Computing Scoring Functions Outline

Grid-based calculation: Introduction
• Precompute “force field” for each term of scoring function
for each conformation of protein (usually only one) Scoring functions
• Molecular mechanics
• Sample force fields at positions of ligand atoms • Empirical functions
Accelerate calculation of scoring function by 100X • Knowledge-based

Searching poses & conformations

Xi p • Systematic search
r
• Molecular dynamics
• Simulated annealing
• Genetic algorithms
• Incremental construction
• Rotamer libraries

[Huey & Morris] Results & Discussion

Systematic Search Systematic Search

Uniform sampling of search space
• Relative position (3)
• Relative orientation (3)
• Rotatable bonds in ligand (n)
• Rotatable bonds in protein (m)
Uniform sampling of search space
The search space • Exhaustive, deterministic
has dimensionality • Quality dependent on granularity of sampling
3 + 3 + rn + rm • Feasible only for low-dimensional problems

Rotations Translations Rotations Translations

tstep rstep tstep rstep

FRED [Yang04] FRED [Yang04]

12
Systematic Search Molecular Mechanics
Uniform sampling of search space Energy minimization:
• Exhaustive, deterministic • Start from a random or specific state
• Quality dependent on granularity of sampling (position, orientation, conformation)
• Feasible only for low-dimensional problems • Move in direction indicated
§ Example: search all rotations – Wigner-D-1 by derivatives of energy function
• Stop when reach
Maximum local minimum
Correlation

Correlation
X
Rotation
Correlation in Rotational
Frequency Correlation
Domain

Molecule 3D Grid Spherical Harmonic

Functions Decompositions
[Marai04]

Simulated Annealing Genetic Algorithm

Monte Carlo search of parameter space: Genetic search of parameter space:
• Start from a random or specific state • Start with a random population of states
(position, orientation, conformation) • Perform random crossovers and mutations
• Make random state changes, accepting up-hill moves to make children
with probability dictated by “temperature” • Select children with highest scores
• Reduce temperature after each move to populate next generation
• Stop after temperature gets very small • Repeat for a number of iterations

AutoDock 2.4 [Morris96] Gold [Jones95], AutoDock 3.0 [Morris98]

Incremental Extension Incremental Extension

Greedy fragment-based construction: Greedy fragment-based construction:
• Partition ligand into fragments • Partition ligand into fragments
• Place base fragment (e.g., with geometric hashing)

FlexX [Rarey96] FlexX [Rarey96]

13
Incremental Extension Rotamer Libraries
Greedy fragment-based construction: Rigid docking of
• Partition ligand into fragments many conformations:
• Place base fragment (e.g., with geometric hashing) • Precompute all low-energy conformations
• Incrementally extend ligand by attaching fragments • Dock each precomputed conformations as rigid bodies

Ligand Predicted
Possible
Conformations
Rigid
Docking

Best Match

FlexX [Rarey96] Protein Glide [Friesner04]

Rigid Docking Outline

This is just like matching binding sites (complement) Introduction
• Can use same methods we used for matching and
indexing point, surface, and/or grid representations Scoring functions
• Molecular mechanics
• Empirical functions
Ligand • Knowledge-based

Searching poses & conformations

• Systematic search
• Molecular dynamics
• Simulated annealing
• Genetic algorithms
• Incremental construction
• Rotamer libraries
Points Surface Grid
Results & Discussion

Discussion References
[Brooks83] Bernhard R. Brooks, Robert E. Bruccoleri, Barry D. Olafson, David J. States, S. Swaminathan, Martin Karplus,
"CHARMM: A program for macromolecular energy, minimization, and dynamics calculations," J. Comp. Chem, 4, 2,
1983, pp. 187-217.
[Eldridge97] M.D. Eldridge, C.W. Murray, T.R. Auton, G.V. Paolini, R.P. Mee, "Empirical scoring functions. I: The development
of a fast empirical scoring function to estimate the binding affinity of ligands in receptor complexes," J. Comput.-
Aided Mol. Des., 11, 1997, pp. 425-445.
[Friesner04] R.A. Friesner, J.L. Banks, R.B. Murphy, T.A. Halgren, J.J. Klicic, D.T. Mainz, M.P. Repasky, E.H. Knoll, M. Shelley,
J.K. Perry, D.E. Shaw, P. Francis, P.S. Shenkin, "Glide: A New Approach for Rapid, Accurate Docking and Scoring.,"
J. Med. Chem, 47, 2004, pp. 1739-1749.
[Gohlke00] H. Gohlke, M. Hendlich, G. Klebe, "Knowledge-based scoring function to predict protein-ligand interactions," J. Mol.

?
Biol., 295, 2000, pp. 337-356.
[Huey & Morris] Ruth Huey and Garrett M. Morris, "Using AutoDock With AutoDockTools: A Tutorial",
http://www.scripps.edu/mb/olson/doc/autodock/UsingAutoDockWithADT.ppt
[Jones97] G. Jones, P. Willett, R.C. Glen, A.R. Leach, R. Taylor, "Development and Validation of a Genetic Algorithm for
Flexible Docking," J. Mol. Biol., 267, 1997, pp. 727-748.
[Marai04] Christopher Marai, "Accommodating Protein Flexibility in Computational Drug Design," Mol Pharmacol, 57, 2, 2004,
p. 213-8.
[Marsden04] Marsden PM, Puvanendrampillai D, Mitchell JBO and Glen RC., "Predicting protein ligand binding affinities: a low
scoring game?", Organic Biomolecular Chemistry, 2, 2004, p. 3267-3273.
[Mitchell99] J.B.O. Mitchell, R. Laskowski, A. Alex, and J.M. Thornton, "BLEEP - potential of mean force describing protein-
ligand interactions: I. Generating potential", J. Comput. Chem., 20, 11, 1999, 1165-1176.
[Morris98] Morris, G. M., Goodsell, D. S., Halliday, R.S., Huey, R., Hart, W. E., Belew, R. K. and Olson, A. J. "Automated
Docking Using a Lamarckian Genetic Algorithm and and Empirical Binding Free Energy Function", J. Computational
Chemistry, 19, 1998, p. 1639-1662.
[Muegge99] I. Muegge, Y.C. Martin, "A general and fast scoring function for protein-ligand interactions: A simplified potential
approach," J. Med. Chem., 42, 1999, pp. 791-804.
[Rarey96] M. Rarey, B. Kramer, T. Lengauer, G. Klebe, "A Fast Flexible Docking Method using an Incremental Construction
Algorithm," Journal of Molecular Biology, 261, 3, 1996, pp. 470-489.

14