Cheminformatics and QSAR

Schematics
Introduction to Chemoinformatics
Simulation and force fields
Energy minimization
Virtual screening
Binding site information
Steps of Drug Designing
Docking and Validation
Ligand and Structure based descriptor
Free Energy Binding
Linear Interaction Energy
Structure Activity Relationships
Validation
Data Analysis
Applications
Summary
 Information technology designed to
Bioinformatics generate and access biological data
and derive information from it

Informatics

Information technology used to

Chemoinformatics design molecular libraries to
interact with identified targets
Related Aspects
Bioinformatics and cheminformatics are generic terms that
encompass the design, creation, organization, management,
retrieval, analysis, dissemination, visualization and use of chemical
and biological information

IT for managing chemical information and solving chemical

problems

Chemistry + Information science + Computer science

Driven by drug discovery research

Chemoinformatics tasks

Manage information about

Chemical properties
Chemical synthesis
Biological effects
Chemoinformatics (Jürgen Bajorath, 2004)

 Chem-, chemi-, or chemo-informatics

 Focus on the information resources needed to optimize
the properties of a ligand to become a drug (Frank
Brown, 1998)

Decision support by computational means

Drug discovery

 Chemical Informatics: the application of information

technology to chemistry (not with a specific focus on drug
discovery)
Frank Brown’s Definition
 The mixing of information resources to transform data
into information and information into knowledge, for the
intended purpose of making decisions faster in the arena
of drug lead identification and optimization.

 Chemoinformatics is the amalgamation or structure of

those chemical information resources to transform data
into vital information and chemical information into
knowledge for the intended purpose of making better
decisions faster in the area of drug lead identification
and organization (Brown FK , 1998)
Bajorath’s Conclusions (Jürgen Bajorath, 2004)

 Boundaries between bioinformatics and chemoinformatics

are fluid

 Both should be closely combined or merged to

significantly impact biotechnology or pharmaceutical
research
Analysis of Chemoinformatics
in the Drug Industry
 The computer is used to analyze the interactions between
the drug and the receptor site and design molecules with
an optimal fit.

 Once targets are developed, libraries of compounds are

screened for activity with one or more relevant assays
using High Throughput Screening.
 HITS

Biochemical target

Test many compounds

Compounds with desired
effect on target = Hit

 OR 
 
Hit directly used Hit provides
as chemical structure for
probe to study chemists to begin
target developing a drug
or other product
Analysis of Cheminformatics
in the Drug Industry
 Hits are then evaluated for binding, potency, selectivity,
and functional activity.

 Seeking to improve:

 Potency
 Absorption
 Distribution
 Metabolism
 Excretion
 Virtual Screening

Rapid computational mining

of 3D molecular databases is
central to generating new
drug leads.

The algorithms must be able

to handle hundreds of
thousands of molecules.
 . 1. Define the target binding
. site points.
..
. 2. Match the distances.
N

F
.
.. 3. Calculate the
H N
N

transformation matrix for

O S

F
. . the orientation.
..
H N
N

O S

4. Dock the molecule.

N

Score the fit.

H N

5.
N

O S

H N
N

O S
 Steps for Drug Designing
 ISIS Draw 2.3
 ChemSketch 8.0. LIGAND PREPARATION
 LigPrep (Schrodinger, 2007)
 Tatutomers, steric isomers and geometry
minimization PROTEIN PREPARATION
 Minimization by MMFF with default setting
 Target protein
 Chains extraction.
DOCKING
 Hydrogens were added via the Maestro interface
(Schrodinger, 2007)
 Removed water molecules
 GLIDE SP
Final preparation was using PPrep (Schrodinger
2007)
 Minimization by OPLS-2005 force field (Polak et
al., 1969) was stopped either after 5000 steps or GLIDE XP
after the energy gradient converged below 0.05
kcal/mol.
 Docking by Glide 4.0 (Halgren, 2004).
 Glide SP (single precision mode). LSBD
 Glide XP (extra precision mode).
Ligand & Structure Based Descriptor (LSBD)
(Schrodinger, 2008)

 The Ligand & Structure-Based Descriptors panel provides a

convenient interface to several Schrödinger programs, which are used
to generate descriptors for a QSAR model.

 The focus is on generating descriptors for a set of ligands that are

docked to a receptor. Three of the programs, Liaison, Prime MM-
GBSA, and the eMBrAcE module of MacroModel, operate on the
ligand and the receptor. The other two, QikProp and Ligparse, operate
on the ligand only.

 The descriptors extracted from Liaison, Prime MM-GBSA, and
eMBrAcE are energetic properties related to ligand binding. QikProp
generates ADME properties, and Ligparse generates structure based
properties such as functional group counts. These descriptors can be
used as input to the model-generation facility in Strike.
 Docked binding site residue

cd ligplot
c://ligplot ligplot file_name.pdb 1 15 B
 Ligplot of Binding Site
http://www.ebi.ac.uk/thornton-srv/databases/cgi-bin/
pdbsum/GetPage.pl?pdbcode=PDBID
QSAR and Drug Design

Compounds + biological activity

QSAR

New compounds with

improved biological activity
Steps in QSAR
 QSAR model (Alam et al., 2008, 2009, 2010, 2011, 2012)
 Quantitative structure-activity relationship (QSAR) is one of the most important
methods in chemometrics, which give information that is useful for drug design and
medicinal chemistry (Marder et al., 2001, Tuppurainen et al., 1999).

 A QSAR equation is a mathematical equation that correlates the biological activity to

a wide variety of physical or chemical parameters (Hansch et al., 2001, Livingstone et
al., 2000).

 (N = 78; R2 = 0.909; S = 0.236; PRESS = 4.840; q2 = 0.888; F = 24.28 )

 We used a more systematic way of variable selection in order of missing value test →
zero test → simple correlation test → genetic algorithm to obtain a set of good
descriptors.
Descriptor examples

 Physical Properties

 MW
 log P (ocanol/water partition)
 bp, mp
 Dipole moment
 solubility
Descriptor examples
 Structural descriptors
 2D
 Atom/Bond counts
 Number non-H atoms
 Number of rotatable bonds
 Number of each functional group
 2C chains, 3C chains, 4C chains, 5C chains, etc.
 Rings and their size

 3D
 Number of accessible conformations
 Surface area
 Potential energy
 Solvation energy
 Water accessible surface area
 Water accessible surface area of all atoms with positive (negative)
partial charge
Pharmacophore Descriptors

 Number of acidic atoms

 Number of basic atoms
 Number of hydrogen bond
donor atoms
 Number of hydrogen bond
acceptor atoms
 Number of hydrophobic
atoms
 Sum of VDW surface areas
of hydrophobic atoms
Lipinski’s Rule of 5 (Lipinski et al., 1997)
 Potential drug candidates should

 Have 5 or fewer H-bond donors (expressed as the sum

of OHs and NHs)
 Have a MW <500
 LogP less than 5
 Have 10 or less H-bond acceptors (expressed as the sum
of Ns and Os)
Pharmacophore and Binding site

Basic idea:
 Appropriate spatial
disposition of a small 6.7
number of functional
4.2-4.7
groups in a molecule is 4.8
5.2
sufficient for achieving a
desired biological effect. 5.1-7.1

 The ensemble formation

will be guided by these
functional groups.
Distances between
binding groups
in Angstroms and
the type of interaction
is searchable
Structural and Pharmacophore Descriptors (Alam et. al., 2012
Data Analysis
Validation of model
 The cross validation analysis performed by using the leave one out
(LOO) method in which one compound removed from the data set
and its activity predicted using the model derived from the rest of the
data points. Prediction error sum of squares (PRESS) is a standard
index to measure the accuracy of a modeling method based on the
cross validation technique. The r²cv was calculated based on the PRESS
and SSY (Sum of squares of deviations of the experimental values from
their mean) using following formula.
Application of QSAR
 Summary

 Chemical Informatics is an evolving field with many facets.

 It will become increasingly important in areas of chemistry outside the
drug industry.
 It will play an increasing role in the developing area of proteomics.

 Energy based prediction model as an efficient tool for generating more

potent and specific inhibitors of receptor site by testing rationally
designed lead compounds based on Drug derivatives.

 Using a combination of topological, electro-topological-state indices,

electronic and thermodynamic descriptors of chemical structures, can
be built robust QSAR model for prediction of cytotoxic activity of any
drug.

 Therefore, these models should facilitate the rational design of novel

derivatives, guide the design of focused libraries based on the skeleton
and facilitate the search for related structures with similar biological
activity from large databases.
 Bibliography
 Bajorath, Jürgen, Ed. Chemoinformatics: Concepts, Methods, and Tools for Drug Discovery.
(Methods in Molecular Biology; 275) Humana, 2004. ISBN 1-58829-261-4
 Brown FK (1998). Chemoinformatics, what it is and how does it impact drug discovery. Annual
Reports in Medicinal Chemistry 33: 375-384.
 Lipinski CA, Lombardo F, Dominy BW, Feeney PJ (1997). Experimental and computational
approaches to estimate the solubility and permeability in drug discovery and development
settings. Advanced Drug Delivery Reviews 23: 3-15.
 Schrodinger LLC. , http://www.schrodinger.com, (accessed: 24. 04.2007).
 Polak E, & Ribiere G (1969). Revue Francaise Inf. Rech. Oper., Serie Rouge. 16-R1, 35-43.
 Halgren TA, Murphy RB, Friesner RA, Beard HS, Frye LL, Pollard WT, Banks JL (2004). Glide: a
new approach for rapid, accurate docking and scoring. 2. Enrichment factors in database
screening J. Med. Chem. 47:1750-1759.
 Lyne PD, Lamb ML & Saeh JC (2006). Accurate Prediction of the Relative Potencies of Members of
a Series of Kinase Inhibitors Using Molecular Docking and MM-GBSA Scoring. J. Med. Chem. 49
4805-4808.
 Marder VJ( 2001). Thrombolytic therapy. Blood Rev. 15:143-157.
 Tuppurainen K (1999). Frontier orbital energies, hydrophobicity and steric factors as physical
qsar descriptors of molecular mutagenicity. A review with a case study: MX compounds,
Chemosphere. 38:3015-3030.
 Hansch C, Kurup A, Garg R & Gao H (2001). Chem-Bioinformatics and QSAR: A Review of QSAR
Lacking Positive Hydrophobic Terms. Chem. Rev. 101:619-672.
 Livingstone DJ (2000). The Characterization of Chemical Structures Using Molecular Properties.
A Survey, J. Chem. Inf. Comput. Sci. 40:195-209.