Updated
Updated · MIT News · Oct 15
MIT Researchers Build IRIS to Trace 6 Cell-Signaling Pathways Across Embryos
Updated
Updated · MIT News · Oct 15

MIT Researchers Build IRIS to Trace 6 Cell-Signaling Pathways Across Embryos

1 articles · Updated · MIT News · Oct 15

Summary

  • Nature Methods published MIT’s Sept. 8 study showing IRIS can infer which signals a cell received at different developmental stages, including in cell types the model had not previously seen.
  • Six major signaling pathways left consistent gene-activity “fingerprints” across cell types, overturning the assumption that each pathway must be mapped separately in every cell type.
  • Thousands of human embryonic stem cells exposed to dozens of pathway combinations trained the neural network, which then correctly tracked signaling in mouse gastrulation cells headed for heart, gut, muscle and spinal cord tissue.
  • Mouse-embryo experiments also confirmed an IRIS prediction for lung-development signaling, suggesting the model could help steer stem cells into desired fates and improve organoids for drug testing and regenerative medicine.

Insights

If cellular signaling leaves a universal signature across species, can this AI tool predict developmental diseases before they even begin to form?
Could a hidden genetic fingerprint discovered by MIT's new AI finally unlock the secret to growing perfect human organs in the lab?
Can an AI time machine that reads a cell's history rewrite the future of regenerative medicine and cure complex respiratory diseases?