Professor David Jones (Head of Bioinformatics Group)

My main research interests are in protein structure prediction and analysis, simulations of protein folding, Hidden Markov Model methods, transmembrane protein analysis, machine learning applications in bioinformatics, de novo protein design methodology, and genome analysis including the application of intelligent software agents. New areas of research include the use of high throughput computing and very large scale machine learning for bioinformatics applications, analysis and prediction of protein disorder, expression data analysis and the analysis and prediction of protein function and protein-protein interactions using deep learning techniques. Over the years I have authored a number of widely-used bioinformatics tools such as PSIPRED, GenTHREADER, MEMSAT and DISOPRED.

Professor David Jones - website

Dr Kevin Bryson (Computational Systems Biology Lab)

Our lab applies computational and mathematical techniques to understand biological systems that generally have clinical relevance.

One area of research is modelling bacteria at the atomic, molecular and population levels. Bacteria are relatively simple and so we can attempt to gain a mechanistic-level understanding of their working. Gaining such an understanding is clinicially important in terms of finding new approaches to tackle drug-resistant strains.

A different area of our research looks at more complex systems where gaining a mechanistic understanding is not viable. In particular we are interested in neurodegenerative diseases such as Alzheimer's, Parkinson's and Huntington's disease. We approach understanding these complex systems by integrating different types of high-throughput data and applying a variety of machine learning and network-based approaches to search for patterns in the data that could suggest different disease mechanisms. Again we try to integrate data both at the molecular level within neurons (e.g. genomics and transcriptomics data) and also at the population level (i.e. whole brain using MRI and EEG data).

Dr Kevin Bryson - website

Dr Daniel Buchan

Benchmarking and applications of machine learning in protein biofinformatics

PSIPRED Website: A complete ground up rebuild of our website and server infrastructure http://bioinf.cs.ucl.ac.uk/psipred_beta/

BioD3: A library of functions for drawing protein annotation diagrams on the web https://github.com/psipred/biod3/

Analystics Automated: A lightweight, workflow management framework for delivering Data Science analyses over the web. https://analyticsautomated.github.io/

Dr Daniel Buchan - web page

Dr Joe Greener

My interests include protein design, protein structure prediction, computational drug discovery, software development and open science. I am currently a research associate working on using sequence information to explore protein structure as part of the ProCovar project. The basis of this is that there is vastly more data on protein sequences than protein structures. By comparing sequences of proteins in the same family we can find covarying positions, and these indicate residues that are in contact in the 3D structure. We will use modern techniques such as synthetic biology and deep learning to extract as much structural information as possible from the available sequence data.

Dr Joe Greener - website

Dr Shaun Kandathil

ProCovar: Methods and applications of covariation in protein sequence alignments.

Proteins play many important roles in the body's cells. This array of functions is made possible by the diverse three-dimensional structures that these molecules adopt. Obtaining structural information about proteins in the 'wet' laboratory is expensive, time-consuming, and doesn't always work. On the other hand, obtaining the sequence of amino acids in a given protein is much easier. Many researchers have therefore taken to trying to predict the three-dimensional structures of proteins computationally. Recently, it has been shown that the accuracy of such predictions can be greatly improved if one knows or can predict beforehand which amino acid pairs will be in close contact in the structure. These contacting residue pairs can be inferred by looking at closely related sequences of a given protein and looking for patterns in the way the sequences change during evolution. My current research is focused on formulating and improving techniques for protein contact prediction. I am using cutting-edge machine learning methods such as deep learning, which have recently been shown to excel at such complex pattern-recognition tasks. I will also be applying similar methods to related problems in protein structure, and protein-protein and protein-nucleic acid interactions. My project is funded by the European Research Council.

Dr Shaun Kandathil - profile

Dr Cen Wan

Project title: BBSRC Flagship Strategic LoLa Drosophila Development Interactome Project

Project description: This project aims to quantify protein dynamics during embryonic development of Drosophila melanogaster. The specific focus of UCL Bioinformatics group is to develop novel machine learning-based approaches for predicting changes in gene function and changes in the protein-protein network. We also aim to develop novel machine learning methods to investigate the new hypotheses are generated across the project as a whole.

Dr Cen Wan - profile

Lewis Moffat

My current work is a part of the ProCovar project funded by the European Research Council. This project covers the usage of covariation based methods in protein structure prediction and analysis. I am focusing using bioinformatics and machine learning techniques to predict protein-RNA interactions leveraging amino acid covariation. My research interests also include the use of deep generative methods to learn a distribution across the protein space.

Lewis Moffat - profile

Completed Projects

  • iSPIDER - An informatics platform for proteomics using Grid-based technologies
  • BioMap - Functional and structural resources for bioinformatics
  • BioRat - Information extraction for biological research
  • BioSapiens - A European virtual institute for genome annotation
  • e-Protein - Fully automated distributed pipeline for large-scale structural and functional annotation
  • Stem cell - Bioinformatics for Stem Cell Transcriptics
  • Geneweaver - Software agent based genome annotation
  • BBSRC Responsive mode grant: New Developments of Large-scale Automatic Protein Function Prediction using Graphical Learning Techniques
  • Elsevier function annotaton and drug discovery
  • How microbiota can shape the effects of drugs in C. elegans-E. coli model system with http://www.cabreirolab.org