Chemical

5. Lipinski

8/12/04

6:37 pm

Page 855

insight review articles

Navigating chemical space

for biology and medicine

Christopher Lipinski1 & Andrew Hopkins2

1

Pfizer Global R&D, Groton Laboratories, Eastern Point Road, Groton, Connecticut 06340, USA (e-mail: christopher_a_lipinski@groton.pfizer.com)

Pfizer Global R&D, Sandwich Laboratories, Ramsgate Road, Sandwich, Kent CT13 9NJ, UK (e-mail: andrew.hopkins@pfizer.com)

2

Despite over a century of applying organic synthesis to the search for drugs, we are still far from even

a cursory examination of the vast number of possible small molecules that could be created. Indeed, a

thorough examination of all ‘chemical space’ is practically impossible. Given this, what are the best strategies

for identifying small molecules that modulate biological targets? And how might such strategies differ,

depending on whether the primary goal is to understand biological systems or to develop potential drugs?

T

he relationship between chemistry, biology and

medicine has been remarkably productive over

the past century, since Paul Ehrlich pioneered the

idea of systematically searching for drugs. By

screening just over 600 synthetic compounds,

Ehrlich discovered arsphenamine (Salvarsan)1, which

greatly improved the treatment of syphilis. Researchers now

routinely screen millions of compounds in the search for

some that are biologically active. Yet even the compound

files of the largest pharmaceutical companies (which typically

contain approximately 106 compounds) offer only a cursory

examination of all the possible organic compounds that

comprise ‘chemical space’ (Box 1). Chemical space is for all

practical purposes infinite and limited only by the chemist’s

imagination.

Not all biologically active compounds have the desired

physicochemical properties to be a drug. A biologically active

compound may be too lipophilic (greasy) to be orally

absorbed, too polar to cross the...