This review presents an AI‑driven tri‑phase framework—transcriptomic, clonal, and structural—for identifying tumor‑reactive CD8⁺ T cells. It critically addresses key bottlenecks: data bias, generative functional blind spots, and clinical validation deficits. The authors propose integrating dynamic physical simulations with multimodal generative AI to enable force‑activated functional prediction, ultimately shifting from static affinity optimization toward programmable, patient‑specific immunotherapy.

The only known supernova remnant of its kind in the Milky Way has an unexpectedly intricate structure: chains of gas knots resembling pearls on a string, each wide enough to fit the solar system multiple times. Using the Gemini North telescope, an international team co-led by Ilaria Caiazzo, assistant professor at the Institute of Science and Technology Austria (ISTA), captured the most detailed view yet of Pa 30, revealing striking new features of the remnant associated with a historical supernova witnessed nearly 850 years ago. The findings were published in The Astrophysical Journal.