Study details
The peer-reviewed study was published in the journal Nature. It grew out of a collaboration between Dr. Efrat Shema’s lab at Weizmann, Hebrew University’s Prof. Guy Ron, and physicians at Kaplan Medical Center in Rehovot. Researchers in France, Germany, Italy, and the United States also participated.
The scientists mapped breast tumors’ epigenetics, the chemical modifications that dictate which genes are turned on or off without altering the genetic code itself. They focused on nucleosomes, the microscopic packages of DNA wound around structural proteins that cells release when they break down.
The team placed blood samples onto specialized glass slides and treated them with fluorescent antibodies engineered to recognize dozens of distinct chemical tags on these protein spools. Using high-resolution, single-molecule fluorescence microscopy, the imaging system scanned millions of these spools one by one. Advanced computational algorithms and artificial intelligence were used to sift through millions of optical signals, isolating the coordinated pattern of a malignant tumor.
The researchers tested the method on blood samples from 77 breast cancer patients who had been followed for at least five years at the Mayo Clinic in Florida, comparing them with samples from 30 healthy women. The trained artificial intelligence models distinguished between the two groups, achieving 95 percent accuracy in detecting early-stage breast cancer, when the tumor was still confined to the breast, measuring less than two centimeters.