Ian Simpson

Dr Ian Simpson is a Principal Investigator in the Patrick Wild Centre and the University of Edinburgh’s Institute for Adaptive and Neural Computation in the School of Informatics.

As the global population grows, so too does the rate of neurological disease – particularly among people in the developing world. Those of us concerned with the start and end of life face an increasing challenge to develop fast and effective ways to discover the underlying mechanisms of these diseases and new ways of designing drugs to treat them.

DATA SCIENCE FOR THE BRAIN
simpson
Comparing the groups of genes that have been linked to different brain diseases can be a very useful way to understand the biology that underlies them and what sets them apart from each other. In this diagram the diseases are represented as numbers around the outside of the circle and the ribbons represent genes “shared” between the diseases. The width of the ribbon is proportional to the number of genes in common between any two diseases. This visualisation reveals complex relationships between these diseases and suggests that for many there is a high degree of commonality in underlying mechanism.

Our research aim is to increase our understanding of the molecular underpinning of neurological diseases and to identify more specific and effective ways in which to treat and prevent them. We believe that in using a sound mechanistic-systems based approach, we will be pinpoint new high quality drug targets, which are cheaper and quicker to produce than they are at the moment.

To this end we are developing statistical, machine learning, modelling and computer science approaches in collaboration with colleagues in the Patrick Wild Centre who are involved in experimental neuroscientific research as well as pharmaceutical companies. We apply these research methods to the study autism spectrum disorders, ADHD and Parkinson’s disease.

Our methodological approaches fall into four main categories:-

  1. Statistical and machine learning based integration of bio-molecular data types e.g. hierarchical Bayesian networks, coDA, ‘graph’ based methods
  2. Digitisation of biology e.g. ontological, graph-based mapping, natural language processing (NLP) methods
  3. Dynamical modeling of biological systems across scales e.g. rule-based-languages, Kappa, BioPepa
  4. Algorithm, software and pipeline development e.g. R/Bioconductor packages, databases, web-applications and services.

Email: Ian.simpson@ed.ac.uk

Full Profile