MIT Whitehead Institute researchers developed a computational approach to decode cell communication. Their model, IRIS, detects unique gene activity fingerprints left by signaling pathways. These fingerprints are consistent across cell types for the same pathway. IRIS can reconstruct signaling histories across many cell types. The model was trained on human embryonic stem cell responses to signaling pathway combinations. IRIS accurately predicted signaling pathway activation in mouse embryos. The approach accelerates stem cell engineering and disease modeling. It also informs therapies by guiding stem cells into specific functional types. The findings were published in Nature Methods. IRIS enables faster decoding of cell communication than experiments alone.

• 6 major signaling pathways
• published in Nature Methods on Sept. 8, 2026