computer-aided drug design/discoverydhmg.amu.edu.pl/uploads/moleculartherapies_lecture_new.pdf ·...

COMPUTER-AIDED

DRUG DESIGN/DISCOVERY

Małgorzata Szeląg, PhDoffice 1.118

[email protected]

Drug design (rational drug design), is the inventive

process of finding new medications based on the

knowledge of a biological target

BASIC FACTS




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




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


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:


donors


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



disqualified


Virtual screening strategies in drug design – methods and applications. EWA BIELSKA, XAVIER LUCAS, ANNA

CZERWONIEC, JOANNA M. KASPRZAK, KATARZYNA H. KAMINSKA, JANUSZ M. BUJNICKI


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/



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


• 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


http://bioinfo3d.cs.tau.ac.il/PharmaGist/


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






- Fragment-based


1. Small fragments

docking

2. Chemical groups

addition






- Fragment-based


1. Small fragments

docking

2. Chemical groups

addition

3. Search for right

conformation






- Fragment-based


1. Binding






- Fragment-based


1. Binding

2. Connecting

De novo






- Fragment-based


1. Starting structure

De novo






- Fragment-based



2. Addition of random

chemical groups






- Fragment-based



2. Addition of random

chemical groups

3. The likelihood

function


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


QSAR

http://wwwcmc.pharm.uu.nl/webcmc/camd/3dqsar.html

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=9986717&dopt=Abstract

• ComFA Analysis




1. Superpose many binding

ligands


QSAR


• ComFA Analysis





ligands

2. Overlay a grid on the

molecular superposition


QSAR


• ComFA Analysis





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


QSAR


• ComFA Analysis




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


QSAR

• ComFA Analysis




Steric bulk favorable

Negative potential favorable


QSAR

• Tool for modeling protein

binding sites

• Tool for finding new

active drugs

• Problems

– Need a superposition of

the molecules


QSAR

reference ligand –

fixed

test ligand – flexible

red H-bond donors

yellow H-bond acceptors

red MTX

green DHF

Ligand Superposition… not easy!


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


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


Comparing Descriptors

http://www.ncbi.nlm.nih.gov/htbin-post/Entrez/query?db=m&form=6&dopt=r&uid=0009834908

Structure-based drug design

Match receptor and ligand

Accuracy & speed

Score the ligands

Receptor Ligands

Compounds libraries

proprietary

ACD, CMC, etc.


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


Energetics of ligand binding


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


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)


Shape matching

1 2

3

Receptor: HIV-1 protease

Active site: Asp groups


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


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


Incremental construction - FlexX

DOCK

FlexX

Schneider et al., Drug Discovery Today.

2002


DOCK vs FlexX

• United protein description: Superposition of receptor

structures.

• Merge similar parts and treat dissimilar parts as

separate alternatives (instances)


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.


FlexE vs FlexX


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


Receptor

Ligands

Scoring functions


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

computer-aided drug design/discoverydhmg.amu.edu.pl/uploads/moleculartherapies_lecture_new.pdf ·...

Documents