The purpose of this group is to bring together researchers, primarily from the biological sciences, computer sciences, and data science, but also from the mathematical and physical sciences and engineering with the aim of identifying new ways of using ML/AI to discover fundamental organisational principles of living systems. 

Explaining the science

In 1952, Alan Turing published a seminal paper that sought to describe organisational principles of living systems from a mathematical view. Living systems exhibit multiple layers of complexity from molecular scales to the whole organisms and ecosystems. Now, both the volume and diversity of experimental data types in biology is growing rapidly, making integrated quantitative analysis across disparate data types a key challenge in modern biological research. 


By bringing together researchers with different expertise this interest group aims to explore the use of machine learning and AI to address fundamental and general biological research questions in a way that would be very difficult to achieve for a single lab or research group.

In particular we aim to:

  • To address the general challenges in using AI for scientific discovery in the life sciences.
  • Provide an organising hub for activity at the Turing, and the wider community of researchers. 
  • To identify and promote collaboration toward grand challenges, and to nucleate larger funding opportunities.

Talking points

How do we build interpretable machine learning models of living systems?

Challenges: Machine learning is a powerful tool for uncovering patterns within data. However, the internal representation of these models may be difficult to interpret.

Example output: Strategies for design and interpretation of machine learning models.

What are the key challenges in combining diverse and multi-scale datasets to build models of living systems?

Challenges: Experimental datasets are often not standardised or interoperable, have sufficient metadata or are not available in formats well suited for large scale machine learning.  

Example output: Case study bringing together different dataset types for integrated quantitative analysis.