We are interested in developing new or improved ways of representing, modelling, and comparing three-dimensional protein and DNA structures in order to help relate molecular structure to biological function. From a computing science point of view, there are two main problems which need to be addressed if we are to make progress towards this "structural bioinformatics" goal. Naturally, we need robust and accurate computational algorithms but we also need flexible and powerful ways of bringing together or integrating information from diverse data sources and databases. These two themes are central to our current research activities.
Current projects include developing protein docking algorithms, modelling conformational change during docking, protein model-building using Constraint Logic Programming, and searching small-molecule 3D databases for drug-like ligands.
Searching Small Molecule 3D Databases using Spherical Polar Correlations (2002-2005)
EPSRC Project Studentship: Andreas Athanasopoulos.
Supervisor: Dave Ritchie.
This project will use our spherical polar Fourier correlation approach to peform fast 3D shape matching on molecular structure databases.
Modelling Conformational Change in Protein Docking Simulations (2001-2004)
EPSRC DTA Studentship: Diana Mustard.
Supervisor: Dave Ritchie.
This project will develop our work on the protein docking problem. The aim here is to explore efficient ways of modelling the conformational changes that often take place when proteins bind together, to help find a tractable way to incorporate structural flexibility into our docking models.
Computer Modelling Studies of Fish Cytokines and their Receptors (1997-2005, part-time)
University Postgraduate Studentship: Antonis Koussounadis.
Supervisors: Dave Ritchie and Chris Secombes.
The aims of this project are to compare and contrast model cytokine and receptor structures from different vertebrate classes to gain a better understanding of the molecular evolution of immune systems.
Molecular Evolution of Novel Combinatorial Proteins. (1998-2001)
BBSRC Industrial CASE Studentship: Martin Swain.
Supervisors: Graham Kemp and Graham Carter (Eclagen Ltd).
This project largely involved the development of a novel Constraint Logic Programming approach to the side-chain placement problem.
Development of a Mediator to Integrate Access to Databases in Molecular Biology (1998-2001)
Research Grant from the BBSRC/EPSRC Joint Programme in Bioinformatics.
Research Associates: Chris Robertson & Nicos Angelopoulos.
Investigators: Graham Kemp, John Fothergill and Peter Gray.
The aims of this project are to develop a mediator to integrate access to heterogeneous, distributed biological databases, to develop a network user-interface to facilitate multi-database querying, and to provide the ability to use constraints for cross-checking data from different sources.
Modelling Protein-Protein Interactions using 6D Fourier Correlation Techniques (1999-2000)
Research Grant from BBSRC.
Research Fellow: Dave Ritchie.
Investigators: Graham Kemp and John Fothergill.
The aims of this project are to develop improved computational methods of modelling the interactions between pairs of globular proteins, building on work done in an earlier project.
Computer Representation of Protein Shape. (1995-1998)
BBSRC Chemicals and Pharmaceuticals Directorate Special Research Studentship: Dave Ritchie.
Supervisors: Graham Kemp and John Fothergill.
In this work we developed parametric surface shape representations and novel computational techniques for rapid matching of similar and complementary surfaces.
Computer Modelling Studies of Major Histocompatibility Complex Molecules (1995-1998)
University Postgraduate Studentship: Tony Brooks.
Supervisors Graham Kemp and John Fothergill.
This work has application to other work in Aberdeen including research into uveitis in the Department of Ophthalmology and Goodpasture's syndrome in the Department of Medicine and Therapeutics. MHC-peptide binding predictions are being used to guide experimental work.
Hex is an interactive protein docking and molecular superposition program written by Dave Ritchie.
This is a program for predicting peptides that are likely to bind to MHC class II molecules, written by Tony Brooks.
This is a main-memory version of the P/FDM object-oriented database system developed by Peter Gray's group.
The European Bioinformatics Institute is establishing a public repository for microarray based gene expression data, named ArrayExpress. Graham Kemp has implemented aeFDM, which is a functional data model implementation of the ArrayExpress schema.
Ensembl is a joint project between EMBL-EBI and the Sanger Centre to develop a software system which produces and maintains automatic annotation on eukaryotic genomes. Graham Kemp has implemented ensemblFDM, which is a functional data model implementation of the Ensembl schema, based on Ensembl's relational tables.
These links are under construction. I'm slowly converting my personal bookmarks (which are a mess!) into something slightly more organised. Here, "bioinformatics" mostly means "structural bioinformatics"...