打破对称BREAK SYMMETRY02 / MEDICINE AI / PROTEIN INTELLIGENCE
Reading the complex
language of life.
We connect structures, states and experiments in a traceable evidence → prediction → discriminating experiment → feedback loop. Models narrow the space; experiments retain the final say.
01Structures & ensembles
02Experimental supervision
03Traceable research workflows
MEDICINE AI / WORKING NOTES
Medicine AI research notes
01
MEDICINE AI · 2026.08.18
↗02Ask which state before asking how strongly it binds
Static structures and a single Kd can flatten the biology that matters. State, time and experimental context belong in the question.
MEDICINE AI · 2026.08.15
↗03A single structure is not the whole protein
From qFit to Dyna-1, information compressed or missing in old experiments may become a new source of supervision for protein states.
MEDICINE AI · 2026.08.10
↗The smallest useful loop in Medicine AI
Progress is not one more model. It is closing an evidence → prediction → discriminating experiment → feedback loop.