UCL Discovery
UCL home » Library Services » Electronic resources » UCL Discovery

The prediction of human oral bioavailability of drugs.

Hills, E.E.; (2008) The prediction of human oral bioavailability of drugs. Doctoral thesis , University of London. Green open access

[thumbnail of U591604.pdf] PDF
Access restricted to UNSPECIFIED

Download (32MB)


The aim of this project was to develop a model for prediction of bioavailability of drug compounds from structural features. Literature bioavailability data was gathered for over 700 compounds. This data was evaluated and assigned a quality score. Prodrugs and drugs known to be actively transported were excluded from the dataset. The factors that affect bioavailability include solubility, permeability and first-pass metabolism. Each of these factors was considered. The first stage was to develop a model to predict solubility and to study the effects of pH on solubility. Solubility was measured in water for a set of 99 compounds selected from the bioavailability dataset, using HPLC as the analytical method. Solubility was also measured in simulated gastric fluid, SGF (pH 1.2), and simulated intestinal fluid (fed state and fasted state), FeSSIF and FaSSIF at pH 6.5. An equation for change in solubility with pH was shown to perform quite well. An Abraham equation was developed to account for solubility enhancement due to the presence of micelles in simulated intestinal fluid. Abraham equations were derived for prediction of diffusion in water and in ethanol using literature data. Diffusion in ethanol was predicted best and was used in place of permeability to classify the bioavailability data according to the Biopharmaceutics Classification System. A compound which is highly metabolised will have low bioavailability. Human plasma clearance data was used as a measure of the degree of metabolism of a drug. A molecular fragment-based approach using PLS statistics was used to develop a model of human plasma clearance of drugs for over 400 drugs. The final stage of this work was to develop a model of bioavailability. The best model used PLS statistics and included chain-based molecular fragments in combination with calculated aqueous solubility, diffusion coefficients (in ethanol) and plasma clearance.

Type: Thesis (Doctoral)
Title: The prediction of human oral bioavailability of drugs.
Identifier: PQ ETD:591604
Open access status: An open access version is available from UCL Discovery
Language: English
Additional information: Thesis digitised by ProQuest. Third party copyright material has been removed from the ethesis
UCL classification: UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Maths and Physical Sciences > Dept of Chemistry
URI: https://discovery.ucl.ac.uk/id/eprint/1444302
Downloads since deposit
Download activity - last month
Download activity - last 12 months
Downloads by country - last 12 months

Archive Staff Only

View Item View Item