Photograph of Douglas Armstrong

J Douglas Armstrong

Professor of Systems Neurobiology Director, Institute for Machine Learning School of Informatics, University of Edinburgh

My group works at the join between machine learning and experimental neuroscience: curating and integrating connectomics data for the Drosophila nervous system, building computational models of the synapse, and extracting behavioural biomarkers from continuous video of freely behaving animals. The common thread is turning messy, high-volume biological measurement into something that can be reasoned over.

I am also Director and Chief Scientific Officer of Actual Analytics Ltd, a University of Edinburgh spin-out working on automated home-cage monitoring for preclinical research, and a Visiting Scientist at Forschungszentrum Jülich.

Research

Virtual Fly Brain and connectomics

Virtual Fly Brain is an interactive atlas and query platform for the Drosophila nervous system, integrating neuroanatomy, connectomes, expression data and the associated literature. Wellcome-funded, with collaborators in Cambridge and at the Sanger Institute.

Synapse proteomics and synaptome architecture

Network and modelling approaches to the molecular composition of the synapse, and how its architecture varies across neuron types, development and disease. Largely with Seth Grant's group and supported by BBSRC.

Automated behavioural analysis and digital biomarkers

Continuous, non-intrusive monitoring of group-housed animals in the home cage, and the analytics needed to turn that signal into standardised endpoints — including motor biomarkers for ageing and neurodegeneration, and safety-pharmacology endpoints that snapshot tests miss.

Publications

Rather than keep a list here that goes stale, the maintained records are:

Google ScholarMost complete, with citations Edinburgh Research ExplorerOutputs, projects and datasets ORCID0000-0002-5397-3864

Teaching and supervision

I supervise PhD students on interdisciplinary projects spanning machine learning and biomedical data. If that fits what you want to do, email me a CV and a short outline of your interests. Please also look at the Informatics Graduate School and the UKRI AI CDT in Biomedical Innovation, which are the usual routes in.

Elsewhere

Virtual Fly BrainDrosophila neuroscience platform Actual AnalyticsHome cage analysis, spin-out Institute for Machine LearningInstitute I direct

Contact

Room 2:30, Informatics Forum 10 Crichton Street, Edinburgh EH8 9AB jda@ed.ac.uk