Note: Some of the graphical elements
of this site are only visible to browsers that support accepted
web standards. The content of
this site is, however, accessible to any browser or Internet device.
The main research focus is on the use of Inductive Logic Programming,
together with vector machine learning for Bioinformatic applications
such as identification of toxins which inhibit enzymes in metabolic
networks, prediction of 3 dimensional structure of molecules in biochemical
reactions for drug development and the missing functions within metabolic
and gene expression networks. Main interdisciplinary collaborations
are via the College Centre for Bioinformatics.
Methods
•
Inductive Logic Programming (ILP)
Induction, abduction, genetic algorithms for searching
•
Bayes' nets used for Metabolic Networks
Hidden nodes, structure & parameter ( eg. reaction rates)
estimation
•
Support Vector Machine Learning
Combine with ILP
•
Support Vector Machine Learning
Combine with ILP
•
Probabilistic ILP
Stochastic logic programs. Used for structure and parameter estimation
Applications
•
Functional Genomics
Inference of gene function from activity
•
Predictive Toxicology
Identification of toxins which inhibit enzymes in metabolic
networks
•
Pharmacophore prediction
3D structure of molecules in biochemical reactions – key
for drug development
•
Biological network completion
Hypothesise missing functions within metabolic and gene expression
networks
