Drug design (rational drug design), is the inventive
process of finding new medications based on the
knowledge of a biological target
BASIC FACTS
Drug design (rational drug design), is the inventive
process of finding new medications based on the
knowledge of a biological target
S
S
O
NH2
O
O O
NH
N N
O
R R
OHOH
NH NH2
NH
O
The drug is most
commonly an
organic small
molecule that
activates or inhibits
the function of a
biomolecule such
as a protein
which in turn results in a therapeutic
benefit to the patient
Drug design (rational drug design), is the inventive
process of finding new medications based on the
knowledge of a biological target
S
S
O
NH2
O
O O
NH
N N
O
R R
OHOH
NH NH2
NH
O
Drug design involves the
design of small molecules
that are complementary in shape and charge to
the biomolecular target
with which they interact
and therefore will bind to it.
Drug design that relies on
the knowledge of the three-
dimensional structure of
the target is known as
structure-based drug
design.
Drug design frequently relies
on computer modeling
techniques - computer-
aided drug design.
Review and approval by Food
& Drug Administration (USA)
Phase III: Confirms effectiveness and monitors
adverse reactions from long-term use in 1,000 to
5,000 patient volunteers.
Phase II: Assesses effectiveness and
looks for side effects in 100 to 500 patient
volunteers.
Phase I: Evaluates safety and dosage
in 20 to 100 healthy human volunteers.
Discovery and preclininal testing:
Compounds are identified and evaluated
in laboratory and animal studies for
safety, biological activity, and formulation.
5,000-10,000
compounds
0 2 4 6 8 10 12 14
Years
16
Source: Tufts Center for the Study of Drug Development
1 compound
approved
Bringing a New Drug to Market
$400M - $800M
~5-10 compounds
enter clinical trials
Clinical Trials Phase I
About 5 years after start of development
10 – 20 healthy human volunteers
Tolerability, Pharmacokinetics, Pharmacodynamics, Excretion
Phase II
100 – 200 patients in the clinic
Dosage finding
Unwanted side effects
Phase III
>200, >2000 patients inside and outside of clinics
Is the drug therapeutically effective?
Rare side effects, Allergic effects
Application for admission
Phase IV
5 years strict surveillance by the authority
Asessment of side effects
Every 5 years the admission must be prolonged
Continuous studies in animals and humans
http://clinicaltrials.gov/ct2/info/understand
Do New Drugs Always Have to
Cost So Much?
No official data from
pharma companies
The molecule is patented for 20 years
During this time only the patent owner can use the patent commercially
(e.g. sell it or license it to others).
But all competitors can do research on this molecule.
After 20 years everybody can use the molecule commercially (e.g. develop
generic drugs)
In order to increase profits (or make the effort of drug development
profitable at all!), time and cost of development of active compounds,
before they enter the clinical trials, have to be reduced.
Patenting
http://www.fda.gov/default.htm
http://www.ema.europa.eu/ema/
The perfect drug
LIPINSKI`S RULE OF FIVE
Poor absorption or permeation are more
likely when :
- There are more than 5 H-bond donors
- There are more than 10 H-bond acceptors
- The molecular weight is over 500 Da
- The log(P) [n-octanol/water] is over 5
- The sums of N`s and O`s is over 10
•Selectivity
•Lesser Toxicity
•Bioavailability
•Slow Clearance
•Reach The Target
•Ease Of Synthesis
•Low Price
•Slow Or No Development Of Resistance
•Stability Upon Storage As Tablet Or Solution
•Pharmacokinetic Parameters
•No Allergies
A perfect drug
(more than just an inhibitor!)
A perfect drug
Virtual screening strategies in drug design – methods and applications. EWA BIELSKA, XAVIER LUCAS, ANNA
CZERWONIEC, JOANNA M. KASPRZAK, KATARZYNA H. KAMINSKA, JANUSZ M. BUJNICKI
Is generally an enzyme/receptor in a pathway and its inhibition leads to either killing a pathogenic organism or
to modify some aspects of metabolism of body that is
functioning normally.
• An ideal target…
– Is essential for the survival of the organism.
– Located at a critical step in the metabolic pathway.
– Makes the organism vulnerable.
– Exhibits low concentration in the organism.
– Is amenable for simple HTS assays.
A perfect drug target
• Absorption
• Distribution
• Metabolism
• Excretion
ADME concerns can be more important than bioactivity!
Most of these properties are difficult to predict from the chemical structure of the compound!
ADME
Traditional Methods of Drug Discovery
compound effect molecular target
Isolation of active compound
from natural sources
or by serendipity (accidentally)
willow
Acetyl salicylic acid Hipokrates, Charles
Frederic Gerhardt, Felix
Hoffman (1897 Friedrich
BayerCo)
Penicillin Ernest Duchesne,
Alexander Fleming (1928),
Howard Walter Florey &
Ernst Boris Chain
Sulfonamides Gerhard Domagk (1935)
Traditional Methods of Drug Discovery
What`s different?
• Drug discovery process begins with a disease (rather than a
treatment)
• Use disease model to pinpoint relevant genetic/biological
components (i.e. possible drug targets)
Modern Methods of Drug Discovery
compound molecular target effect
Bioactive molecules identified, designed or optimized by CADD methods
(Computer–aided drug design)
Structure Target Method
Aldose reductase inhibitors 3D database searching
Diabetes
HIV protease inhibitorsAIDS
3D database searching
Structure based design
Structure based design Carbonic anhydrase
inhibitor (Dorzolamide)Glaucoma
AnticoagulantsThrombin inhibitors
Combinatorial docking
de novo design
2) Van Drie JH, Lajiness MS, Drug Discovery Today 1998, 3, 274-283
3) Drugs, New, Perspect. 1995, 8, 237.
4) Bohm HJ et al, J. Comput.-Aided Mol. Des. 1999, 13, 51-56
N
R3 Cl
R2
R1
N N
O
R R
OHOH
S
S
O
NH2
O
O O
NH
NH NH2
NH
O
Examples of Success
Where are we?
Preliminary trials
Preclinical trials Phase I Phase II Phase III FDA/EMA Phase IV
2-6 years
drug drug
discovery design
What do we need?
DATA
TOOLS
RESULTS
Success rate Experimental approach
High-throughput screening
400 000 chemical compounds
300 selected compounds IC50 < 300μM
85 compounds IC50 < 100μM
0.021 % success rate
In silico approach
Virtual screening
235 000 chemical compounds
365 selected compounds
127 compounds IC50 < 100μM
34.8% success rate
Doman et al., J Med CheM. 2002
Compound
databases,
Microbial broths,
Plant extracts,
Combinatorial
Libraries
3-D ligand
Databases
Docking
Linking or
Binding
Receptor-Ligand
Complex
Random
screening,
synthesis
Lead molecule
3-D QSAR
Targetidentification
3-D structure by
Crystallography,
NMR, electron
microscopy or
Homology Modeling
Re-design
to improve
affinity,
specificity etc.
Testing
How Bioinformatics Can Help?
Drug design
knowledge-based method
Ligand info No ligand info
Structure-based
drug design
De novo
design
Ligand-based
drug design -
C o m p u t e r – a i d e d d r u g d e s i g n
( C A D D )
Konwn receptor structure
Unknown
receptor
structure
Ligand-based drug design
LIPINSKI`S RULE OF FIVE
Poor absorption or permeation
are more likely when:
- There are more than 5 H-bond
donors
- There are more than 10 H-bond
acceptors
- The mol.wt is over 500 Da
- The log(P) [n-octanol/water] is
over 5
- The sums of N`s and O`s is over
10
(L)ADME(T)
Liberation
Absorption
Distribution
Metabolism
Excretion
Toxicity
Drug-like
Similarity with known
drugs
Database search with
molecular descriptors
Similarity in drug design
Taxol
LIPINSKI’S RULE OF FIVE
Poor absorption or permeation are
more likely when:
- There are more than 5 H-bond
donors
- There are more than 10 H-bond
acceptors
- The mol.wt is over 500 Da
- The log(P) [n-octanol/water] is
over 5
- The sums of N`s and O`s is over
10
Similarity in drug design
Ligand-based drug design
disqualified
Similarity in drug design
Virtual screening strategies in drug design – methods and applications. EWA BIELSKA, XAVIER LUCAS, ANNA
CZERWONIEC, JOANNA M. KASPRZAK, KATARZYNA H. KAMINSKA, JANUSZ M. BUJNICKI
Ligand-based drug design
Find your fingerprint
A B
A
B
Find your fingerprint
1 1 0 1 1 0 1 0
1 1 0 1 0 0 0 0
A B
A
B
• Tools
OpenBabel (Python) http://openbabel.org/wiki/Main_Page
ZINC Search http://zinc.docking.org/search/structure
Phase Shape (Schrödinger) http://www.schrodinger.com/products/14/34/
DRAGON http://www.vcclab.org/lab/edragon/
Similarity in drug design
Ligand-based drug design
A pharmacophore is an abstract
description of molecular features
which are necessary for molecular
recognition of a ligand by a biological
macromolecule.
Typical pharmacophore features include: hydrophobic centroids, aromatic rings,
hydrogen bond acceptors or donor, cations, and
anions.
These pharmacophoric points may be located
on the ligand itself or may be projected points
presumed to be located in the receptor.
Pharmacophores Ligand-based drug design
Zanamivir (Relenza) Oseltamivir (Tamiflu)
HBA
HBD
ring
Pharmacophores Ligand-based drug design
Zanamivir (Relenza) Oseltamivir (Tamiflu)
HBA
HBD
ring
Pharmacophores Ligand-based drug design
• Tools
Pharmer http://zincpharmer.csb.pitt.edu/
PharmaGist http://bioinfo3d.cs.tau.ac.il/PharmaGist/
Catalyst (Discovery Studio, Accelrys) http://accelrys.com/
Molplex http://molmaker.molplex.com/pharma
Pharmacophores Ligand-based drug design
http://bioinfo3d.cs.tau.ac.il/PharmaGist/
Pharmacophores Ligand-based drug design
De novo Ligand-based drug design
• Strategy- Receptor structure
- Binding site model
- Best matching of ligands into binding site
• Approach- Incremental construction
- Fragment-based
- Stochastic optimisation
1. Small fragments
docking
De novo Ligand-based drug design
• Strategy- Receptor structure
- Binding site model
- Best matching of ligands into binding site
• Approach- Incremental construction
- Fragment-based
- Stochastic optimisation
1. Small fragments
docking
2. Chemical groups
addition
De novo Ligand-based drug design
• Strategy- Receptor structure
- Binding site model
- Best matching of ligands into binding site
• Approach- Incremental construction
- Fragment-based
- Stochastic optimisation
1. Small fragments
docking
2. Chemical groups
addition
3. Search for right
conformation
De novo Ligand-based drug design
• Strategy- Receptor structure
- Binding site model
- Best matching of ligands into binding site
• Approach- Incremental construction
- Fragment-based
- Stochastic optimisation
1. Small fragments
docking
2. Chemical groups
addition
3. Search for right
conformation
De novo Ligand-based drug design
• Strategy- Receptor structure
- Binding site model
- Best matching of ligands into binding site
• Approach- Incremental construction
- Fragment-based
- Stochastic optimisation
1. Binding
De novo Ligand-based drug design
• Strategy- Receptor structure
- Binding site model
- Best matching of ligands into binding site
• Approach- Incremental construction
- Fragment-based
- Stochastic optimisation
1. Binding
De novo Ligand-based drug design
• Strategy- Receptor structure
- Binding site model
- Best matching of ligands into binding site
• Approach- Incremental construction
- Fragment-based
- Stochastic optimisation
1. Binding
2. Connecting
De novo
Ligand-based drug design
• Strategy- Receptor structure
- Binding site model
- Best matching of ligands into binding site
• Approach- Incremental construction
- Fragment-based
- Stochastic optimisation
1. Starting structure
De novo
Ligand-based drug design
• Strategy- Receptor structure
- Binding site model
- Best matching of ligands into binding site
• Approach- Incremental construction
- Fragment-based
- Stochastic optimisation
1. Starting structure
2. Addition of random
chemical groups
De novo Ligand-based drug design
• Strategy- Receptor structure
- Binding site model
- Best matching of ligands into binding site
• Approach- Incremental construction
- Fragment-based
- Stochastic optimisation
1. Starting structure
2. Addition of random
chemical groups
3. The likelihood
function
Ligand-based drug design
QSAR Quantitative structure-
activity relationship
• Strategy- Relationship between similarity and activity
- Prediction of compound activity based on its physical and chemical properties
SAR QSAR:
Similar compounds posses similar
activity
Quantitative model of relationship
between structure and activity
• ComFA Analysis
• Approach to find out
parts of a ligand that are
important for function
Böhm et al. 1999
Ligand-based drug design
QSAR
• ComFA Analysis
• Approach to find out
parts of a ligand that are
important for function
1. Superpose many binding
ligands
Ligand-based drug design
QSAR
• ComFA Analysis
• Approach to find out
parts of a ligand that are
important for function
1. Superpose many binding
ligands
2. Overlay a grid on the
molecular superposition
Ligand-based drug design
QSAR
• ComFA Analysis
• Approach to find out
parts of a ligand that are
important for function
1. Superpose many binding
ligands
2. Overlay a grid on the
molecular superposition
3. Compute field properties at
grid points for each
molecule
• Steric field of a carbon
probe
• Electrostatic field
Ligand-based drug design
QSAR
• ComFA Analysis
• Approach to find out
parts of a ligand that are
important for function
4. Solve the linear regression
problem
1 1 1
,
,
Act field
Act field
field is the value of field for
molecule at grid point
parameters to be trained
fields are steric, charge etc.
ij ij
i j
k kij kij
i j
kij
i
i
k j
Ligand-based drug design
QSAR
• ComFA Analysis
• Approach to find out
parts of a ligand that are
important for function
Steric bulk favorable
Negative potential favorable
Ligand-based drug design
QSAR
• Tool for modeling protein
binding sites
• Tool for finding new
active drugs
• Problems
– Need a superposition of
the molecules
Ligand-based drug design
QSAR
reference ligand –
fixed
test ligand – flexible
red H-bond donors
yellow H-bond acceptors
red MTX
green DHF
Ligand Superposition… not easy!
Ligand-based drug design
QSAR
babaN
aaN
baNbNaN
baNbaT
and both in 1set bits ofnumber ),(
in 1set bits ofnumber )(
),()()(
),(),(
tcoefficien Tanimoto
•structur
•hashed fingerprints
•PPP pairs / triangles
00101110101010110001 Typical 2D and 3D descriptors:
Advantages:
• simple compare functions
• can be combined easily
• good performance
Disadvantages:
• exact matches
• overall topology weakly represented
• conformation dependenceN
NH+
N
N
NH2
NH2
N
O
NH
O
O
OO
Ligand-based drug design Molecular Simi larity Descriptors
Ligand-based drug design
Descriptors
• Strategy:– calculate a match between subtrees
optimizing similarity
• Matching Properties:– a match maps two subtrees onto each
other
– each node occurs in at most one match
– matching is topology-maintaining
– matched subtrees should be balancedin size
• Algorithms:– split-search (divide & conquer)
– match-search (dynamic programming)
M.Rarey and S.Dixon, J.Comput.-Aided.Mol.Des. 12 (1998), pp 471
Ligand-based drug design
Comparing Descriptors
Structure-based drug design
Match receptor and ligand
Accuracy & speed
Score the ligands
Receptor Ligands
Compounds libraries
proprietary
ACD, CMC, etc.
Structure-based drug design
Protein-ligand docking
Receptor
Ligand
Affinity: DG = DH -TDS
Displaced H2O Bound and
associated H2O
Upon complex formation:
• water molecules are released
• receptor and ligand loose degrees of freedom
• interactions between ligand and receptor
complication: mutual compensation of enthalpy and entropy
Structure-based drug design
Energetics of ligand binding
Structure-based drug design
Docking algorithms
„Rigid body docking”
Structural modeling
Screening algorithms
„Flexible body docking„
Sampling across entire range of positional,
orientational and conformational possibilities.
Various methods have been developed:
• Fast shape matching (Dock, Eudock)
• Incremental construction (FlexX, Hammerhead)
• (Lamarckian) Genetic algorithm (Gold, Autodock3.0)
• Simulated annealing (Autodock2.4)
• Monte Carlo simulations (MCDock, QXP)
• Distance geometry (Dockit)
Usually RMSD < 2 Å is considered acceptable
Structure-based drug design
Docking methodology
DOCK works in 5 steps:
• Step 1 Start with crystal coordinates of target receptor
• Step 2 Generate molecular surface for receptor
• Step 3 Generate spheres to fill the active site of the
receptor: The spheres become potential locations for
ligand atoms
• Step 4 Matching: Sphere centers are then matched to
the ligand atoms, to determine possible orientations for
the ligand
• Step 5 Scoring: Find the top scoring orientation
DOCK (I. D. Kuntz, UCSF)
Structure-based drug design
Shape matching
1 2
3
Receptor: HIV-1 protease
Active site: Asp groups
Structure-based drug design
Shape matching
4 5
• Three scoring schemes:
Shape scoring, Electrostatic scoring, and Force-field scoring
• Image 5 is a comparison of the top scoring orientation of
thioketal with the orientation found in the crystal structure
Structure-based drug design
Shape matching
Modeling receptor ligand interactions:
- Receptor interaction surface from crystallographic
information etc.
Approximation by a finite set of interaction centers.
Fragmentation of ligand into base fragments.
Place ligands into active site by matching interaction
centers (first triples than line matching for pairs).
Reduction of number of solutions by clash test and
clustering.
Link base fragments in compliance with a torsional
database or a force field.
O
N
H H H
H
S
NH2
O
S
N
O
H
H
O
N
H
N
O
S
O
N
N
H
O
S
H.J. Böhm, J. Comput. Aided Molec. Des. 8, 623-632 (1994)
M.D. Miller, R.P. Sheridan, S.K. Kearsley, J. Med. Chem. 1999, 42, 1505-1514
Structure-based drug design
Incremental construction - FlexX
DOCK
FlexX
Schneider et al., Drug Discovery Today.
2002
Structure-based drug design
DOCK vs FlexX
• United protein description: Superposition of receptor
structures.
• Merge similar parts and treat dissimilar parts as
separate alternatives (instances)
Structure-based drug design
Receptor flexibility- FlexE
1. .........
2. .........
3. .........
: .........
10. ...
FlexE Integrated scoring / ranking
for all structure combinations
All ligands are docked into
the united protein structure.
The individual ranking list are
merged and resorted by score.
1. .........
2. .........
3. .........
: .........
10. ...
1. ........
2. ........
3. ........
4. ........
5. .. 1. ........
2. ........
3. ........
4. ........
5. ..
1. ........
2. ........
3. ........
4. ........
5. ..
FlexX All ligands are sequentially
docked into all structures.
Structure-based drug design
FlexE vs FlexX
Structure-based drug design
Genetic algorithms
PLANTS
Protein-Ligand ANT
System
htt
p:/
/ww
w.t
cd
.un
i-ko
nsta
nz.d
e/r
esearc
h/p
lants
.php
Genetic algorithms belong to the larger class of evolutionary algorithms (EA),
which generate solutions to optimization problems using techniques inspired by
natural evolution, such as inheritance, mutation, selection, and crossover.
Ant colony optimization algorithm
is inspired by the behaviour of
real ants finding a shortest path
between their nest and a food
source.
Protein-ligand interactions
- Disulfide bridges
- Hydrogen bonds
- Hydrophobic interactions
- van der Waals interactions
• IC50 – biological measure of activity
Km - substrate concentration for half of Vmax of reaction
• Kd – termondynamic measure of ligand affinity
Koff - Kinetic dissociation constant
Kon - Kinetic association constant
IC50
1 + [S]/Km Ki =
koff [R]x[L]
kon [RL] Kd = =
Protein-ligand interactions
Thermodynamics of interactions
• Gibbs energy – thermodynamic potential
H - enthalpy
S - entropy
• ΔG < 0 condition for spontaneous reaction
Receptor
in solution
ligand
in a solution system
ΔG= ΔH – T(ΔS)
Scoring functions
Scoring functions
are used during docking for optimization of ligand orientation and
conformation for docked ligands to estimate affinity relative to
other compounds
H.J. Böhm, J. Comput. Aided Molec. Des. 8, 623-632 (1994)
M.D. Miller, R.P. Sheridan, S.K. Kearsley, J. Med. Chem. 1999, 42, 1505-1514
Various criteria for the quality of a docking function:
• find the correct binding mode out of alternative docking
solutions
• rank related ligands with respect to their binding affinity
• select (weak) inhibitors from a large database of inactive
compounds
• be fast and error tolerant
The inaccuracy of functions used to estimate the affinity between receptor
and ligand is considered to be the major weakness of docking programs.
Scoring functions
C. Bissantz, G. Folkers, D. Rognan, J. Med. Chem. 2000, 43, 4759-4767
PS Charifson, JJ Corkery, MA Muecko, WP Walters, J. Med. Chem. 1999, 42, 5100-5109
force field based methods
Dock, Gold, FlexX
separate contributions from hydrogen bonds, ionic and lipophilic
interactions, clashes and entropy (nr of rotatable bonds)
potentials of mean force (PMF)
Drug Score
description of observed interatomic distances and/or frequencies
implying that these describe favorable/unfavorable interactions
consensus scoring
combination of multiple scoring functions increases hit rates by reducing
the number of false positives
two stage ranking
first filter to limit the number of docked conformations,
second filter to reject false positives
Scoring functions
0
5
10
15
20
25
30
Gold FlexX Dock
Hit
Ra
te %
Gol
d +
Fle
xX
Gol
d
PMF
+ F
lexX
PMF
or
Fle
xX
PMF
+ F
lexX
PMF
or
Fle
xX
C. Bissantz, G. Folkers, D. Rognan, J. Med. Chem. 2000, 43, 4759-4767
Docking algorithms: Gold, FlexX and Dock,
Scoring algorithms: Gold, FlexX and PMF
Data set: 990 random compounds and 10 known ligands
Target: Thymidine Kinase
Hit Rate: % of known ligands among top 5% scorers
Hit rates do not strongly dependent on docking tool used
Hit rates significantly improved by consensus scoring
High number of false positives
Benchmarking of docking programs
and scoring functions
• water
• cofactors
• alternative conformations
Remove:
• missing H atoms
• missing structural elements
• information about atom types and charges
Add:
• binding site
• important interactions between water or ions
Identify:
Docking - step-by-step Receptor preparation
Tools
Chimera DockPrep (UCSF)
http://www.cgl.ucsf.edu/chimera/download.html
Maestro Protein Preparation Wizard (Schrödinger)
http://www.schrodinger.com/products/14/12/
• H atoms
• information about atom type and charge
Add:
• tautomers
• stereoisomers
Create:
• degree of freedom
• geometry
Optimization
Docking - step-by-step Ligands library preparation
Tools
Chimera DockPrep (UCSF) http://www.cgl.ucsf.edu/chimera/download.html
OpenEye OMEGA
http://www.eyesopen.com/omega
Maestro LigPrep Wizard (Schrödinger) https://www.schrodinger.com/products/14/10/
http://zinc.docking.org/
Docking - step-by-step Ligands library preparation
Virtual screening The same methods but with big libraries
Receptor
Ligands
Searching algorithms
Virtual screening The same methods but with big libraries
Receptor
Ligands
Scoring functions
Virtual screening The same methods but with big libraries
Receptor
Ligands
1 2 3 4 5 6
Ranking of top ligands
http://www.click2drug.org/
EXAMPLES
Vitrual screening for inhibitors
of RNA-dependent RNA polymerase NS5
of Dengue virus
• RNA-dependent RNA polymerase NS5of Dengue virus is required for the formation of the viral mRNA cap structure.
Ewa Bielska
NS5-MTaza
mRNA
cap
Background
• Dengue fever also known as breakbone fever, is an
infectious tropical disease caused by the dengue virus.
• Dengue is transmitted by several species of mosquito.
Ewa Bielska
Background
• Dengue has become a global problem since the Second
World War and is endemic in more than 110 countries.
• Apart from eliminating the mosquitoes, work is ongoing
on a vaccine, as well as medication targeted directly at
the virus.
Ewa Bielska
Step-by-step
Crystal structure of receptor with SAH and SAM molecules
Ligands conformation in active site
Ewa Bielska
Looking for targets?
ORPHANETThe portal for rare diseases and orphan drugs.
Orphanet`s aim is to help improve the
diagnosis, care and treatment of patients with
rare diseases.
„There is no disease so rare
that it does not deserve attention”
THE END