Immunity By Design

Oxford University
  • 0.0 (0)
  • Education

From Cells to Systems Through Human and Machine Intelligence. Modern immunology stands at a transformative turning point. Emerging AI and experimental technologies are rapidly reshaping how we do and understand immunology. But this acceleration comes at a cost. A widening gap seems to emerge: those who adapt alongside these changes prove more effective at translating ideas into impact, while even the most brilliant concepts can be consumed by complexity and strategy. And in the middle, a central challenge remains: how do we move beyond pattern recognition to genuine mechanistic understanding? How do we use AI not as a black box for prediction, but as a tool to decode underlying biological principles? 'Immunity by Design: From Cells to Systems Through Human and Machine Intelligence', explores these questions with the pioneers in science an innovation as well as representatives of policymakers. Hosted by Professor Hashem Koohy -- a leading scientist with a research focus on systems immunology  the podcast brings together researchers, technology developers, entrepreneurs, and investors reimagining immune science for the digital age. Each episode explores how data, AI, and systems immunology are reshaping discovery: how researchers combine spatial and single-cell technologies, immune-repertoire profiling, and in-silico innovations to reveal disease mechanisms and inform therapeutic development; how experimental design can enable interpretable, mechanism driven AI; and how we trace the path from data to insight, from model to medicine, and from idea to impact.

Episodes

  1. Jun 2

    Beyond Black Boxes: Decoding T Cell Recognition with Interpretable AI

    Professor David Gfeller discusses interpretable AI, TCR specificity profiles, cross-reactivity, chain pairing, and why understanding T cell recognition may matter more than black-box prediction. In this episode of Immunity by Design, Prof Hashem Koohy speaks with Professor David Gfeller (University of Lausanne & Ludwig Institute for Cancer Research) about the emerging rules governing T cell recognition and how interpretable AI may reshape computational immunology. The conversation explores three recent studies from David Gfeller’s group focused on TCR specificity profiles (TSPs), probabilistic modelling of TCR recognition, and the role of alpha-beta chain pairing in T cell specificity prediction. Together, these studies challenge several widely held assumptions in the field, including the dominance of black-box sequence prediction models, the centrality of paired TCR sequencing for prediction performance, and the underappreciated role of V and J gene usage in shaping antigen recognition. Key topics discussed include: Why sequence-based TCR prediction struggles to generalize across unseen epitopes o The biological and statistical foundations of TCR specificity profiles (TSPs) o TEMPO: an interpretable probabilistic framework for TCR epitope prediction o The importance of baseline repertoires in modelling T cell specificity o Structural insights into V/J gene usage and TCR recognition o Why alpha-beta chain pairing may contribute less predictive information than expected o Cross-reactivity, neoantigen recognition, and off-target toxicity prediction o The translational implications for TCR-based immunotherapies o Open science, preprints, and the future of AI-driven immunology research o The importance of interpretability in the era of increasingly powerful AI systems The episode also explores broader questions around multidisciplinary research culture, scientific training, and how computational and experimental scientists can collaborate more effectively. Music: “Minimal” by paulyudin, sourced from Pixabay and used under the Pixabay Content License (royalty-free for commercial use).

    1h 2m

  2. May 22

    The Hidden Biology of Crohn’s Fistula

    Dr Agne Antanaviciute joins Prof Hashem Koohy to discuss how spatial omics and single-cell technologies uncovered the hidden biology of Crohn’s fistula and revealed fibroblasts as key drivers of disease. Crohn’s fistula is one of the most severe and understudied complications of inflammatory bowel disease, causing chronic pain, recurrent infection, and profoundly reduced quality of life. Yet for decades, little was understood about the cellular and molecular mechanisms driving these destructive lesions. In this episode of Immunity by Design, Prof Hashem Koohy is joined by Dr Agne Antanviciute, a corresponding author of the recent Nature paper “Spatial Fibroblast Niches Define Crohn’s Fistula”, to explore how cutting-edge spatial omics and single-cell technologies are reshaping our understanding of chronic intestinal inflammation. The conversation explores: - what Crohn’s fistula actually is and why it is so devastating for patients - how multimodal technologies including single-cell RNA sequencing, spatial transcriptomics, Xenium, Visium, and collagen imaging were integrated to build the largest molecular atlas of Crohn’s fistula to date - the discovery of specialised “FAS fibroblasts” that appear to drive tissue tunnelling, fibrosis, and chronic inflammation - how pathological wound healing and immune–stromal interactions may underpin disease progression - why these findings could have implications far beyond Crohn’s disease, including fibrosis and chronic inflammatory disorders in other organs - the future role of AI and computational biology in interpreting increasingly complex multimodal datasets The episode also reflects on multidisciplinary science, research culture, collaboration, scientific failure, and the changing future of biomedical research in the era of AI. Paper discussed: “Spatial Fibroblast Niches Define Crohn’s Fistula”, Nature (2025) Host: Prof Hashem Koohy, University of Oxford Guest: Dr Agne Antanaviciute Music: “Minimal” by paulyudin, sourced from Pixabay and used under the Pixabay Content License (royalty-free for commercial use).

    56 min

  3. Apr 14

    Can AI Solve Snakebite? With Dr Timothy Jenkins

    Prof Hashem Koohy is joined by Dr Timothy Jenkins to explore how AlphaFold and experimental technologies are transforming antibody discovery, enabling molecules that neutralise multiple snake venom toxins. Snakebite envenoming remains one of the most neglected global health challenges, causing over 100,000 deaths each year, yet antivenom design has changed little in more than a century. In this episode of Immunity by Design, Dr Timothy Jenkins (head of Centre for Translational Protein Design, at the Technical University of Denmark) joins us to discuss a new paradigm for antibody discovery at the interface of AI and biology. We explore his team’s recent work (bioRxiv: 2026.01.19.700436, link below), where they develop an integrated pipeline combining phage display, yeast display, deep sequencing, and structure prediction (AlphaFold3) to identify polyspecific nanobodies -- molecules capable of neutralizing multiple snake venom toxins simultaneously. This represents a fundamental shift from traditional approaches, moving from empirical discovery toward more predictive, design-driven strategies. The conversation unpacks not only how these technologies were brought together, but why this moment -- driven by advances in machine learning and structural biology-- makes such integration possible. We discuss the strengths and limitations of current tools (including AlphaFold), the concept of polyspecificity as a design principle, and how multi-objective optimisation (affinity, solubility) can be achieved without compromising function. Looking beyond snakebite, we explore how this pipeline could be applied across immunology and biomedicine— from viral escape and pandemic preparedness to cancer and biosecurity, highlighting the importance of democratising AI tools for tackling neglected and resource-limited diseases. Finally, we reflect on the human side of science: building multidisciplinary teams, fostering positive research culture, and navigating non-linear scientific careers at the interface of disciplines. bioRxiv: 2026.01.19.700436: https://www.biorxiv.org/content/10.64898/2026.01.19.700436v1.abstract Music credit: Intro and outro features music from the Isabella Stewart Gardner Museum, Boston by Colin Carr.

    48 min

  4. Mar 25

    Decoding T Cell Recognition at Scale by Dr Stephanie Gaglione

    Dr Stephanie Gaglione discusses scalable technologies to map T cell receptor antigen interactions, enabling large datasets that could transform how we study immune recognition and build computational models of immunity. In this episode of 'Immunity by Design', Dr Hashem Koohy speaks with Dr Stephanie Gaglione about new experimental approaches that could transform how we study T cell recognition. At the heart of adaptive immunity lies the interaction between T cell receptors (TCRs) and antigenic peptides presented by MHC molecules. However, mapping these interactions at scale has remained one of the central challenges in immunology.  Dr Gaglione discusses her work on scalable strategies for reconstructing and screening large numbers of TCRs, including the TCRAFT approach developed during her PhD with Michael Birnbaum at MIT. By enabling the reconstruction of thousands of TCRs from sequence data and integrating functional screening technologies such as RAPTR, these methods open the door to generating high-quality datasets linking TCR sequences to antigen specificity.  The conversation explores the technological hurdles of many-to-many TCR antigen screening, the importance of accessible experimental tools for building community-scale datasets, and how such data could accelerate the development of computational models that predict immune recognition. The episode also touches on applications in autoimmune disease, including vitiligo, and the broader implications for immunology, cancer research, and data driven discovery.

    47 min
  • 4 Episodes

About

From Cells to Systems Through Human and Machine Intelligence. Modern immunology stands at a transformative turning point. Emerging AI and experimental technologies are rapidly reshaping how we do and understand immunology. But this acceleration comes at a cost. A widening gap seems to emerge: those who adapt alongside these changes prove more effective at translating ideas into impact, while even the most brilliant concepts can be consumed by complexity and strategy. And in the middle, a central challenge remains: how do we move beyond pattern recognition to genuine mechanistic understanding? How do we use AI not as a black box for prediction, but as a tool to decode underlying biological principles? 'Immunity by Design: From Cells to Systems Through Human and Machine Intelligence', explores these questions with the pioneers in science an innovation as well as representatives of policymakers. Hosted by Professor Hashem Koohy -- a leading scientist with a research focus on systems immunology  the podcast brings together researchers, technology developers, entrepreneurs, and investors reimagining immune science for the digital age. Each episode explores how data, AI, and systems immunology are reshaping discovery: how researchers combine spatial and single-cell technologies, immune-repertoire profiling, and in-silico innovations to reveal disease mechanisms and inform therapeutic development; how experimental design can enable interpretable, mechanism driven AI; and how we trace the path from data to insight, from model to medicine, and from idea to impact.

Information

  • Creator
    Oxford University
  • Years Active
    2K
  • Episodes
    4
  • Rating
    Clean
  • Copyright
    © Oxford University; the media items are released with a Creative Commons licence
  • Show Website
    Immunity By Design