Making sense of cells: The role of AI in advancing biomedical and computational biology

Ghent (Belgium), June 12, 2024 – It was a full office at the BioLizard head office, with diverse life science professionals meeting there to attend the latest Flanders AI Academy (VAIA) Highway Session on AI in the Life Sciences.

At this event, four speakers presented their cutting-edge work that merges biology and artificial intelligence to extract new scientific knowledge and actionable insights.

First, Prof. Kris Laukens (University of Antwerp & ImmuneWatch) outlined how immunopeptidomics, single cell sequencing and T cell receptor repertoire sequencing can be leveraged to obtain unique, rich insights into the complex interactions between peptide antigens and components of the adaptive immune system. He also shared concrete case studies from ImmuneWatch to demonstrate how new computational tools and approaches can improve the development processes of smart vaccines, personalised therapies and precision diagnostic applications.

Next, Prof. Yvan Saeys (VIB & Ghent University) gave a scientifically and visually beautiful talk about how single cell spatial omics are revolutionizing our understanding of tissue dynamics. He walked us through some of the novel machine learning techniques, including cell-cell communication models, that are being developed in his lab to understand these "cellular symphonies", and how they open up exciting possibilities for defining new and potentially more targeted biomarkers, grounding the discussion in cancer studies as use cases. 

Prof. Stein Aerts (VIB & KU Leuven) then demonstrated how his laboratory seamlessly blends experimental and computational biology techniques in order to decipher the genomic regulatory code. Walking us through this approach that mixes single cell ATAC-seq experimental assays, deep learning implementations for enhancer modelling, and AI-driven design of synthetic enhancers, Prof. Aerts shared both exciting scientific discoveries and concrete tools that his lab has developed to drive forward the research field.

Last but not least, our own Dr. Andrea Del Cortona shared different strategies for overcoming the challenge of integrating high-dimensional, multi-omics data, and guided us through 2 case studies in which BioLizard used Bio|Mx - our fully cutomizable multi-omics data analysis, exploration & visualization platform - to help biotech and biopharma clients overcome different data challenges to answer scientific questions. The take-home message? The approach must fit the data, and a mix of data science and biological know-how is the key to success.

Finally, at the dynamic panel discussion moderated by Dr. Alex Cloherty (BioLizard), the expert speakers discussed topics such as... 

•  the need for education that trains biologists to blend wet-lab experimentation and computational / data science approaches,
• the importance of developing explainable AI models to gain richer scientific insights and boost their uptake in industry,
• and the importance of shifting the norms in the funding of scientific research, in order to reward scientists for validating data and replicating findings - which could ultimately improve the quality of public data and thereby our ability to leverage it using AI.

Thank you to the knowledgeable speakers & attendees for your engagement, as well as the teams at VAIA and Flanders AI Research Program for the organization.