QSAR

Quantitative Structure Activity Relationship is a

multivariate mathematical relationship between a set of 1D/2D/3D physicochemical properties (descriptors) and any other property, say biological activity.

The QSAR relationship is expressed as a mathematical equation. Analysis of statistical relationships between structure and various properties is required in understanding how chemical structure and biological activity could relate to each other.

Equation of the line passing through the origin then

Y = mx Equation of line whose Y intercept is C Y = mx + C m is slope

QSAR
Addresses two questions:
What features of a molecule affect its activity?

What can be modified to enhance properties?

First correlation: Shape and Ionisation

Albert showed that Acridines become progressively more active (bacteriostatic activity) as they become more highly ionized.
H
H N H

Ionisation
N

N H
N

H N H

N H H

H N

N H H

Bacteriostatic activity

N H H

The attention was dragged towards the biological activity and correlated the structural parameters to the activity for different classes of molecules such as pyridines, quinolines, benzoquinolines and phenanthridines.

The influence of Physical organic chemistry:

By the influence of ionization on drug activity, physical chemists were developing parameters to quantify the electronic effects of substituents on chemical reactions and ionic equilibria. The Hammett and Taft correlated the chemical activity of the compound with the variation of substituents at different positions.

For a set of compounds, the equation is

Log (KX/KH) = x
KX is the rate constant of the derivative the rate of different reactions depends on the electron release or withdrawal of the substituents

KH is the rate constant for the parent

x refers

to the electronic effect of the substituents relative to hydrogen

Development: QSAR equation

Three groups tried to get biological Hammett equation. 1. Hansen Tried on sets of substituted benzoic acids, phenyl substituted penicillin-G and chloramphenicol analogues Gave very disappointing results There are number of reasons for failure

But it marked an important stage in the development of QSAR theory, and provided a foundation for later thinking

2. Zahradnik
Tried to develop a biological Hammett equation for more fundamental principles Replaced the parameters and with and Chose his standard biological reaction as the toxicity of alcohols to mice (whereas Hammetts standard reaction was ionization of benzoic acids) Approach failed for 2 reasons: (a) Research was concentrated on single parameter,

(b) Too close to Hammett recipe, looked for the simplicity of a linear relationship

3. Hansch Approach to QSAR

Worked on different sets of molecules Activity correlation using multi-parameters to get QSAR (Approached with a different idea) Found that the relative lipophilic character of the substituents was an important determinant of the activity The idea - for any particular receptor, some optimum value of log(P) would be found to correspond to the maximum probability of reaching the receptor in a given time.

The simplest way to express such an idea is that log (1/C) is parabolically dependent on log (P).
Thus the extremely useful general Hansch equation was proposed.

Log(1/C) = k1(logP) k2(logP)2 + k3 + k4

Hammetts sigma.

The QSAR
Any QSAR can be expressed in its most general form by equation, Biological activity = f(physicochemical and/or structural parameter)

Overall objective

To find parameters (from experiment or theory) that when substituted into one of the many forms of the equation along with biological activity for a series of molecules should give a statistical significant correlation.

Y = C1X1 + C2X2 + C3X3 +

QSAR objectives can be divided broadly as

Diagnosis of mechanism (drug transport to site of action) Prediction of activity (interpolative or extrapolative) Classification (highly active, active, inactive) Optimization (activity by steric, electronic, hydrophobicity etc.) Refinement of synthetic targets Reduction and replacement of animals (due to the ban of toxic chemicals on the animals, FRAME-For Replacement of Animals with Modified Experiments)

How to generate QSAR?

This procedure consists of the following 10 steps: 1. Identify the training set 2. Enter biological activity data 3. Generate conformations 4. Calculate descriptors 6. Generate a QSAR equation

5. Explore the data

7. Validate the equation

8. Analyze the equation 9. Save the QSAR equation 10. Predict activity

Application Examples
QSAR of Sweeteners
Examined a family of artificial sweeteners Evaluated data using - Used 33 QSAR descriptors - Principle components analysis - Genetic algorithms - Molecular field analysis (MFA)

Predictive model established to test new sulfamates

Application Examples
Structure-Activity Relationships in drug design
Investigators were studying Curacin A Antimiotic agent Lipid that interferes with tublin assembly QSAR analysis showed completely different patterns than colchicine like drugs 65 descriptors used from QSAR+

Application Examples
Toxicology
Examination of toxicity of a series of organic compounds
IC50 measured by toxicity to Daphnio magna, the water flea Used basic QSAR techniques Identified partition coefficient and hydrogen bonding as important properties

Few Examples from Many Successes of RDD

NOROXIN, antibacterial agent - Kyorin Pharmaceutical
Norfloxacin: QSAR techniques were used in its development. Following its discovery, 6-fluoroquinolones became a major class of antibiotics.

Koga, H. et al. J. Med. Chem. 1980, 23, 1358-1363.

Few Examples from Many Successes of RDD

COZAAR for Hypertension treatment DuPont, Merck
Losartan: Angiotensin II receptor antagonist developed via molecular modeling of a lead compound from patent literature and QSAR

Duncia, J. V. et al. J. Med. Chem. 1990, 33, 1312-1329; Duncia, J. V. et al. Med. Res. Rev. 1992, 12, 149-191.

Few Examples from Many Successes of RDD

ARICEPT for Alzheimers disease - Eisai

Donepezil: Acetylcholinesterase inhibitor developed via QSAR, molecular shape analysis, docking

Kawakami, Y. et al. Bioorg. Med. Chem. 1996, 4, 1429-1446.