Last week on The Genetics Podcast, Patrick Short sat down with Michelle Teng, CEO and co-founder of Etcembly, to explore her work at the intersection of immunotherapy and machine learning. The discussion ranged from breakthroughs in cancer treatment to Michelle’s deeply personal motivation for advancing rare disease therapies. In case you missed it, here's a quick recap of the episode:
Michelle traced the evolution of immunotherapy from its speculative beginnings in the 1980s to its role in modern cancer care. By "releasing the brakes" on the immune system, these therapies empower T-cells to identify and eliminate cancer cells. However, despite significant breakthroughs, Michelle emphasized ongoing challenges, including treatment-related toxicity, tumor escape mechanisms, and inconsistent patient responses. These obstacles highlight the critical need for improved biomarkers, refined patient selection strategies, and a more comprehensive understanding of immune system biology.
At Etcembly, Michelle and her team are pioneering a new approach to immunotherapy by studying long-term cancer survivors. Through the Long-Term Survivors Study, they analyze immune repertoires—specifically T-cell and B-cell receptors—of individuals whose immune systems have successfully controlled cancer. Using EMILY, their proprietary machine learning algorithm, they aim to identify immune receptor sequences that contribute to durable cancer defense.
Michelle compared this process to large language models like ChatGPT. Instead of predicting words, EMILY deciphers the "language" of immune receptor sequences to uncover patterns. By mapping known and unknown receptors, Etcembly is building a library of immune responses that can inform the development of new therapies.
Etcembly’s approach bridges computational prediction with experimental validation. Immune receptors identified by EMILY are tested in the lab for tumor-killing activity. These results feed back into the model, improving its predictive power. This iterative process is what is accelerating the discovery and engineering of next-generation immunotherapies, offering hope for patients with difficult-to-treat cancers.
Michelle’s commitment to precision medicine extends beyond cancer research. She also founded a company called SynaptixBio after her daughter was diagnosed with a rare leukodystrophy caused by a mutation in the TUBB4A gene. Now, SynaptixBio is developing RNA-based therapies targeting the underlying genetic defect.
Looking ahead, Michelle envisions combining different therapeutic approaches, such as antibody-drug conjugates (ADCs) with T-cell receptor therapies, to create more multifaceted cancer treatments. For now, her work at Etcembly demonstrates the potential of computational tools to transform our understanding of immune responses and advance precision medicine. For a deeper dive into Michelle’s research and personal journey, make sure to check out the full episode of The Genetics Podcast.
Listen now: