Jealott’s Hill International Research Centre, Bracknell, UK
The programme for the Spring meeting of the UK QSAR & CI Group covered topics of current interest in pharmaceutical and agrochemical research. Martin Battersby gave a short introduction to Syngenta and its research activities. John Delaney (Syngenta) discussed the use of HT in vitro and in vivo screens as a source agrochemical leads. The talks by Mike Abraham (UCL) , Derek Reynolds (GSK) and Eric Clarke (Syngenta) included a common theme of physicochemical parameter determination while Alexander Alex (Pfizer) and Paul Bamborough (GSK) shared a theme of ligand binding, scoring and QSAR. Tudor Oprea (AstraZeneca) gave a talk relating to drug likeness and library design.
- Mike Abraham (University College London, UK)
Prediction of physical chemical and biological processes from structure using LFER methodology [Slides] [Abstract]
- Alexander Alex (Pfizer, Sandwich, UK)
Prediction of protein-ligand binding energies: Thoughts and experiences [Abstract]
- Paul Bamborough (GlaxoSmithKline, Stevenage, UK)
Experiences with Docking, Scoring and QSAR [Slides] [Abstract]
- Martin Battersby (Syngenta, UK)
Discovery and Syngenta
- Eric Clarke (Syngenta, Jealott’s Hill, UK)
Bioavailability: An agrochemical research perspective [Slides] [Abstract]
- John Delaney (Syngenta, Jealott’s Hill, UK)
What have we learnt about using HT screening as a source of agrochemical leads? [Slides] [Abstract]
- Tudor Oprea (AstraZeneca, Sweden)
The design of lead like combinatorial libraries [Slides] [Abstract]
- Derek Reynolds (GlaxoSmithKline, Stevenage, UK)
Rapid physicochemical profiling [Slides] [Abstract]
Presentation: Mike Abraham
Prediction of physical chemical and biological processes from structure using LFER methodology
University College London, UK [Slides]
Transport properties of nonelectrolytes can be satisfactorily correlated and predicted by multiple linear free energy relationships that use a number of specific solute descriptors as the independent variables. These descriptors are: E the solute excess molar refraction, S the solute polarizability/ dipolarity, A the solute hydrogen-bond acidity, B the solute hydrogen-bond basicity, and V the solute McGowan volume. Recently, we have calculated these descriptors from structure, using a group contribution scheme, so that predictions from structure can now be made. The general LFER equation is:
SP = c + e.E + s.S + a.A + b.B + v.V (1)
Here SP is a solute property in a given system, such as logP for water/octanol partition. Equations based on eq.(1) have been published for water/solvent partitions, solubility in water, and blood/brain distribution, and further examples will be shown, particularly that for human intestinal absorption.
In addition to the use of the general LFER for correlation and prediction, examples will be given of its use for the easy visualisation, and for the chemical interpretation, of various processes.
Presentation: Alexander Alex
Prediction of protein-ligand binding energies: Thoughts and experiences
Pfizer, Sandwich, UK
Virtual screening is now widely considered to be a useful addition to the computational methods applicable to the Drug Discovery process.1 In this presentation, we will focus on our experiences with current docking and scoring methodology applied to a variety of data sets. We will also present results generated using the empirical scoring potential BLEEP developed by Mitchell et al.2,3 Comparisons will be made between current scoring functions and other methods for predictions of binding energies such as QSAR and FEP studies. We will discuss in general the strengths and weaknesses we have identified in current scoring functions, and we will highlight possible future developments in this exciting area of research.
1. H.-J. Boehm, M. Boehringer, D. Bur, H. Gmuender, W. Huber, W. Klaus, D. Kostrewa, H. Kuehne, T. Luebbers, N. Meunier-Keller, F. Mueller, J. Med. Chem. 2000, 43, 2664.
2. J.B.O. Mitchell, R.A. Laskowski, A. Alex, J.M.Thornton, J. Comput. Chem. 1999, 20, 1165.
3. J.B.O. Mitchell, R.A. Laskowski, A. Alex, M.J. Forester, J.M. Thornton, J. Comput. Chem. 1999, 20, 1177.
Presentation: Paul Bamborough
Experiences with Docking, Scoring and QSAR
GlaxoSmithKline, Stevenage, UK [Slides]
Docking compounds into a protein structure is an established method for screening databases to find new chemical leads or for predicting a compound’s binding mode in the absence of an experimental structure. A related problem is the ranking of members of a single chemical family in order of their experimental binding affinity. A variety of rapid computational scoring functions that attempt to do this have been proposed. Here a group of P38 kinase inhibitors from a GSK chemistry programme have been docked using the Cambridge Crystallographic Data Centre’s Gold algorithm. Different methods are evaluated for their ability to predict the binding affinity of the compounds in this series.
Presentation: Martin Battersby
Discovery and Syngenta
Presentation: Eric Clarke
Bioavailability: An agrochemical research perspective
Syngenta, Jealott’s Hill, UK [Slides]
The ability of a compound to express activity in a given biological screen relates initially to its availability from screening media and ultimately to its concentration and interactions at a target site within a screen enzyme, organism or species. In an ideal world bioavailability would be assessed via direct measurement of concentration profile from time of application to expression of activity in vitro or in vivo; but in reality such concentrations are seldom, if ever, known with confidence. In this talk the concepts of bioavailability being a composite of mobility and stability related properties; and activity within and across screens assessed in terms of a compound’s relative ‘potency-mobility-stability’ balance are outlined. Measurements of relevance to compound mobility and stability in biological systems made at Jealott’s Hill are highlighted. Mobility related properties include octanol-water distribution coefficients (log Doct), acid-base dissociation constants (pKa), aqueous solubility (log Sw) and volatility. Stability related properties include reactivity in the presence of water, thiols or oxidants with and without catalysis by esterases, glutathione transferases and iron(III)porphyrins; and solid and solution phase photochemistry. The main focus of this presentation is the mobility of agrochemicals. Current & developing methodologies to measure mobility related properties have been applied to a diverse range of ~50 herbicides, fungicides and insecticides used in commercial products to illustrate the range of values found in effective agrochemicals. This set of compounds has also been used to compare methods of estimation for these properties, which include those attributed to Leo, Yalkowsky, and Abraham. Expanding this work, mobility related property ranges for ~ 500 compounds cited in the Pesticide Manual are summarised. Finally, the bioavailability guidelines for agrochemicals deduced from this work are compared with those proposed by Briggs (Agrevo, 1997), Tice (Rohm & Hass, 2001) and those established for drugs by Lipinski et al (Pfizer, 1997).
Presentation: John Delaney
What have we learnt about using HT screening as a source of agrochemical leads?
Syngenta, Jealott’s Hill, UK [Slides]
After several years of using high throughput screening as a source of leads, we have begun to form some conclusions about the factors that affect the number of leads produced. This talk will focus on the quality of chemical input, applying automation and standardisation to the analysis of screen hits, and the effect of screen cycle times on lead follow up.
Presentation: Tudor Oprea
The design of lead like combinatorial libraries
AstraZeneca, Sweden [Slides]
To be considered for further development, lead structures should display the following properties: (1) simple chemical features, amenable for chemistry optimization; (2) membership to an established SAR series; (3) favorable patent situation; (4) good absorption, distribution, metabolism and excretion (ADME) properties.
There are two distinct categories of leads: those that lack any therapeutic use (i.e., “pure” leads), and those that are marketed drugs themselves, but have been altered to yield novel drugs. We have previously analyzed the design of leadlike combinatorial libraries starting from 18 lead and drug pairs of structures (S.J. Teague et al., Angew. Chem. Int. Ed. 1999, 38, 3743-3748). Here, we report results based on an extended dataset of 96 lead-drug pairs, of which 62 are “pure leads”, and 75 are “pure drugs” (i.e., not used as leads).
We examined the following properties: MW (molecular weight), CMR (the calculated molecular refractivity), RNG (the number of rings), RTB (the number of rotatable bonds), the number of hydrogen bond donors (HDO) and acceptors (HAC), the calculated negative logarithm of the n-octanol/water partition (CLogP), the calculated negative logarithm of the distribution coefficient at pH 7.4 (LogD74), the Daylight-fingerprint druglike score (DFPS) and the Property and pharmacophore features score (PPFS). The findings indicate that the process of optimizing a lead into a drug results in more complex structures. This information should be used in the design of novel combinatorial libraries that are aimed at lead discovery.
Presentation: Derek Reynolds
Rapid physicochemical profiling
GlaxoSmithKline, Stevenage, UK [Slides]
Understanding the physicochemical properties of molecules early in the discovery process can help to ensure that development candidates are selected with adequate solubility and appropriate absorption and distribution in-vivo. High-throughput methods are now available for the routine measurement of octanol logP, solubility, and pKa. A very simple and rapid alternative approach to lipophilicity measurement has been devised based on retention in generic gradient HPLC. Values of CHI (Chromatographic Hydrophobicity Index) can be correlated with molecular properties using the Abraham solvation equation. This forms the basis of a convenient experimental method for determination of the molecular descriptors S (dipolarity/polarisability), A (hydrogen bond acidity), and B (hydrogen bond basicity). Reliable experimental molecular descriptors for ‘drug-like’ molecules are invaluable for investigating transport mechanisms in biological systems.