The devil is still in the details--driving early drug discovery forward with biophysical experimental methods |
| |
Authors: | Lundqvist Tomas |
| |
Affiliation: | AstraZeneca Structural Chemistry Laboratory, Global Science and Information, AstraZeneca R&D, M?lndal S-431 83, Sweden. tomas.lundqvist@astrazeneca.com |
| |
Abstract: | This review comments on some recent trends and insights in the field of lead identification and optimization with a bias toward the increased use of biophysical methods, particularly in combination with three-dimensional structural information. While high-throughput screening, combinatorial chemistry and, most recently, in silico virtual screening techniques have made well-resourced but only partially successful attempts to meet the challenge of identifying new drug candidates by playing 'the large numbers game', another group of technologies are now approaching the same challenge from what might be considered the opposite extreme. The common strategy of these technologies is to focus on a smaller set of low-molecular-weight compounds whose interactions with a target are characterized with the aid of sensitive assays, most often high-quality biophysical techniques such as biosensors, calorimetry, nuclear magnetic resonance spectroscopy and X-ray crystallography. The advantages of such an approach include more optimal and chemically attractive starting points, immediate access to reliable measurements of binding properties, the mapping of ligand interactions on the atomic level and, most importantly, a greater control of experimental errors at the initial stages of drug discovery where compounds are either discovered or lost. When correctly supported, this more careful approach appears to deliver quality leads, even for the so-called 'difficult' targets. As these techniques are complementary to traditional methods, companies should be less hesitant to invest in them. The biophysical methods that are used to drive this approach have made something of a return to drug discovery after having been discarded for being too slow, too expensive or too old-fashioned by the over-optimistic supporters of high-throughput and statistical/computational in silico methods. |
| |
Keywords: | |
本文献已被 PubMed 等数据库收录! |
|