The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence) Using AI to Map the Human Immune System w/ Jabran Zahid - #485
May 20, 2021
Jabran Zahid, a Senior Researcher at Microsoft Research, dives into the fascinating world of mapping the human immune system using AI. With a unique background in astrophysics, he discusses the Antigen Map Project and its adaptation during the COVID-19 pandemic. Jabran shares insights into the complexities of T cell development and the challenges faced in using machine learning for immunology. He highlights the importance of model interpretability and the progress made towards developing FDA-approved diagnostic tools for enhanced health understanding.
AI Snips
Chapters
Transcript
Episode notes
From Astrophysics to Immunology
- Jabran Zahid, an astrophysicist, joined Microsoft Research to work on the Antigen Map Project.
- This project aims to map the human immune system, intrigued by its potential impact.
Decoding the Immune System
- The Antigen Map Project maps T-cells and antigens to understand the immune system's memory.
- Sequencing T-cell receptors helps decode past exposures and diagnose active diseases.
T-Cell Generation
- T-cell production is a quasi-random process, creating naive T-cells without specific targets.
- Upon encountering a pathogen, these naive cells proliferate into memory cells, retaining immunological memory.