Autumn Meeting 2000

Charnwood, Loughborough


  • Paul E. Blower (LeadScope)
    Software for exploring large sets of drug discovery data [Abstract]
  • Rob Brown (Molecular Simulations Inc, UK)
    Recursive Partitioning [Slides]
  • Mark Cronin (Liverpool John Moores University, UK)
    Toxicity Prediction [Slides]
  • Andy Davis (AstraZeneca R & D Charnwood, UK)
    The Physical Organic Chemistry of Drug-Phospholipid Interactions and their Role in Pharmacological Action [Abstract]
  • Gavin Harper (GlaxoWellcome, UK)
    QSAR’s using HTS data [Slides] [Abstract]
  • Pete Kenny (AstraZeneca R & D Alderly Park, UK)
    Hydrogen Bonding [Slides] [Abstract]
  • Mark Wenlock (AstraZeneca R & D Charnwood, UK)
    Physical properties of marketed and development oral drugs [Abstract]


Presentation: Paul E. Blower
Software for exploring large sets of drug discovery data


Modern approaches to drug discovery have dramatically increased the speed and quantity of compounds that are made and tested for potential potency. We have developed decision support software to assist pharmaceutical researchers in mining knowledge from the huge volumes of data now being generated by these high-speed techniques. The software organizes the chemical data by structural features familiar to pharmaceutical scientists and provides graphical representations and dynamic querying capabilities that make it easier to interpret the complex data in large structure-activity datasets. This paper will illustrates several techniques for exploring an HTS dataset, testing hypotheses about structure activity relationships, and selecting compounds for further experiments.

Presentation: Rob Brown
Recursive Partitioning

Molecular Simulations Inc, UK [Slides]


Presentation: Mark Cronin
Toxicity Prediction

Liverpool John Moores University, UK [Slides]


Presentation: Andy Davis
The Physical Organic Chemistry of Drug-Phospholipid Interactions and their Role in Pharmacological Action

AstraZeneca R & D Charnwood, UK

Hydrophobicity is important in controlling the absorption, distribution, metabolism, elimination of drugs. For many years this has been modelled by measuring solvent-water partitioning, and n-octanol-water partition coefficients being the most used. This simple physico-chemical system has been widely succesful in guiding the drug design process. The success is quite surprising, as isotropic solvents such as n-octanol are very poor models of the ordered phospholipid bilayer. This talk will focus upon the role of hydrophobicity in controlling drug-membrane interactions, with examples from physicochemical measurements, in-vivo drug distribution studies and QSAR’s in the design of new agents for treatment chronic obstructive pulmonary disease.

Presentation: Gavin Harper
QSAR’s using HTS data

GlaxoWellcome, UK [Slides]

This talk will focus on using statistical techniques, and particularly Recursive Partitioning (RP), to predict biological activity by analysing HTS data.

We begin by describing the general algorithm underlying recursive partitioning (RP). We then proceed to describe a particular version of recursive partitioning (Rusinko et al., 1999). This algorithm has been used to analyse HTS data, and may be used to model the activity of further compounds.

An example will be shown of the application of recursive partitioning to a GlaxoWellcome enzyme assay, and its performance will be compared to that of another algorithm, kernel discrimination (Harper, 1999). This will highlight some of the advantages, and some of the limitations of recursive partitioning.

We will conclude by discussing in a more general setting the use of QSAR techniques for HTS analysis, pointing out some of the potential pitfalls, and encouraging the intelligent integration of QSAR techniques within a broader corporate screening strategy.


Rusinko et al. (1999), J. Chem. Inf. Comput. Sci. 39, 1017-1026.

Harper (1999), DPhil thesis – University of Oxford.

Presentation: Pete Kenny
Hydrogen Bonding

AstraZeneca R & D Alderly Park, UK [Slides]

The use of computer-aided methods to predict the toxicity of drugs is described. These methods can assist in the identification of toxic compounds early in the drug development process. Thus, there is potential for these methods to be combined with combinatorial synthesis and library design. Quantitative structure-activity relationships allow for the prediction of individual endpoints, usually for restricted groups of compounds. Expert systems for toxicity prediction are based on a number of methodologies, each with its own strengths and weaknesses. The relative merit of each individual technique and methodology is described. However, more toxicity data are required both to produce and to validate expert systems. Potential sources of new data include the use of high-throughput screening and microarrays for toxicology.

Presentation: Mark Wenlock
Physical properties of marketed and development oral drugs

AstraZeneca R & D Charnwood, UK

The process of drug development applies rigorous selection pressures, and consequently compounds which have made it to the market, even though as a whole they are very structurally diverse, will generally have favorable physiochemical properties with respect to absorption and metabolism. However, it could be argued that drugs which have successfully made it to the market may not be truly representative of compounds presently in development.

The talk will discuss whether the physiochemical properties of drugs currently in development vary significantly from those of marketed oral drugs. In addition, the distribution of measured LogD(7.4) and aqueous solubility data, on approximately 400 oral marketed drugs, will be presented.