Dr Ian Sillitoe

I am responsible for the technical aspect of CATH. This generally involves maintaining and developing both the front-end interfaces (internal and external web pages and webservices) and back-end code library and databases.

As an undergraduate I studied Chemistry for a 4 year Masters degree at Sheffield University between 1994-1998. I found the topic of organic macromolecular structure and reactions most interesting (specifically protein structure and enzyme catalysis) and this, combined with my dissertation project to create an online resource to teach NMR to undergraduate students, led me to start a Bioinformatics PhD in protein structure/classification here in Prof. Christine Orengo's CATH group at UCL (1998-2002).

After my PhD, I stayed on in the CATH group as a postdoctoral researcher in order to consolidate and extend the work brought up in my PhD and also to help maintain and develop the CATH database (2002-2004). After a couple of years, I decided to have a break from academia and took up a position as the technical director of a commercial web company (2004-2006). A few years later, I decided to have a break from commerce and come back into academia (2006-current)!

During my PhD, I looked at the classification, analysis and recognition of protein structures using the CATH protein structure classification database. I measured structural similarity by comparing contact maps which are defined as points of contact (<8A) between amino acid residues within a protein. By examining structures that were known to be related, it was possible to identify inter-residue contacts that were highly conserved during the process of evolution.

Part of this project involved deriving a protocol for generating accurate multiple structural alignments and 3D templates based on consensus contact patterns found in these alignments (see figure above). Templates, were generated for all homologous superfamilies in CATH to create a library of unique and identifying 'fingerprint' patterns.

These templates were applied to the recognition of models generated at an early stage of ab initio protein structure prediction. Scanning these early models against a library of templates describing conserved contacts allowed the most likely superfamilies to be indentified. I then wrote an algorithm that performed fold recognition using only a limited set of contacts with the purpose of application to the early stages of experimental NMR structure determination.

The multiple structural alignments were also used to generate a library of hidden markov models (HMMs). These structure-based sequence profiles were thoroughly benchmarked using a strict dataset of remote homologues and added sensitivity to existing sequence scans.

I still remember the days when I had time to play piano/keyboard in various soul/blues/jazz bands - now I only get wheeled out for the occasional wedding, bar mitzvah, etc. I also remember the days when I actually used to be quite good at tennis, football and squash. I still play when I get a chance.

I do still go along to ULU Jitsu as often as I can (I was a instructor there from 2005-2007). Any new people that turn up in the lab will probably get hassled mercilessly until they come along to at least one training session (they usually keep coming after that).