Cancer immunotherapy response and dual gene deletion biomarkers in CHD1 and MAP3K7 How tumor gene deletion patterns reveal powerful new immunotherapy biomarkers for predicting cancer treatment success Learn how dual gene loss could guide precision cancer medicine and personalized cancer treatment decisions

What You'll Learn:

• How dual deletion of CHD1 and MAP3K7 functions as a potential biomarker for predicting cancer immunotherapy response
• Why this dual gene loss appears in roughly 8–12% of prostate tumors and 3–5% of urothelial cancers based on TCGA data
• What a retrospective UCSF 2023 study of 109 ICI-treated prostate cancer patients suggests about improved objective response rates in the dual-deletion group (38% vs 11% overall, requiring independent validation)
• How mouse xenograft models lacking both genes showed about a 70% reduction in tumor volume after anti-PD-1 treatment, compared with less than 20% for single-gene knockouts
• What these findings mean for using CHD1 and MAP3K7 status to stratify patients and personalize immunotherapy strategies
• Key caveats, including the need for larger prospective trials and independent validation before clinical implementation
• How dual gene deletion fits into the broader landscape of cancer genetics, tumor microenvironment, and biomarkers for immune checkpoint inhibitor response

Cancer immunotherapy biomarkers and CHD1/MAP3K7 deletion in predicting immunotherapy response New insights into dual-gene deletion, tumor response to immunotherapy, and emerging immunotherapy success predictors Understand how CHD1 and MAP3K7 co-deletion could guide personalized cancer treatment and precision oncology decisions

What You'll Learn:

• How CHD1 and MAP3K7 co-deletion functions as a potential cancer immunotherapy biomarker to predict which patients may benefit most from treatment
• Why CHD1+MAP3K7 co-deletion appears in roughly 8% of primary and 15% of metastatic prostate tumors, and how that prevalence shapes clinical and research priorities
• What preclinical mouse data reveal about a 5-fold increase in CD8+ T-cell infiltration and a 6-fold rise in STING pathway transcripts in dual-deleted tumors
• How a Johns Hopkins pembrolizumab cohort (n=57) showed a 33% objective response rate in dual-deleted tumors versus 7% in non-deleted tumors, and why this signal still requires independent validation
• Practical ways these findings could inform biomarker development, patient stratification, and immunotherapy trial design in prostate cancer and other solid tumors
• How CHD1/MAP3K7 deletion research integrates with existing biomarkers for cancer treatment, including PD-L1, MSI, and TMB, in predicting immunotherapy response
• Key methodological and translational challenges that must be addressed before CHD1/MAP3K7 co-deletion can be used as a routine immunotherapy success predictor in the clinic

Foam physics and artificial intelligence — how everyday bubbles reveal the hidden science of learning and complex systems in nature. This episode uniquely connects foam physics explained with the science of learning in AI, showing how materials that learn and the physics of everyday materials mirror what is artificial intelligence doing under the hood. Listen to discover how AI inspired by nature may uncover a universal logic linking foam bubbles science, artificial intelligence and physics, and living cells.

What You'll Learn:

• How typical wet foam can pack an astonishing 10^8–10^9 bubbles per liter and why that matters for understanding complex systems in nature.
• Why foams were once modeled like glassy, frozen materials—and what new simulations reveal about their constant microscopic motion.
• What a T1 rearrangement is, how a single event lowers surface energy by about 10^-12 joules, and why that tiny change is a powerful unit of ‘learning-like’ behavior.
• How more than 10^4 T1 events per second in a fist-sized foam sample create a dynamic, self-optimizing structure that still keeps its overall shape.
• How the same mathematics that describes bubble rearrangements also underpins how AI learns in neural networks and other machine-learning models.
• What is artificial intelligence in this context, and how AI inspired by nature can borrow ideas from foam physics and the physics of everyday materials.
• Why the science of learning might be a shared principle across materials that learn, artificial intelligence systems, and living cells.
• How these insights could shape future research in artificial intelligence and physics, from smarter materials to new learning algorithms.