Mad-cow disease occurs when an unruly protein called a prion causes healthy proteins in cattle brains to misfold. The same is true for the human versions of mad cow—“variant” Creutzfeldt-Jakob disease, which is contracted from beef, and the naturally occurring “sporadic” form. But until Kelvin Lee unleashed a new style of protein analysis, diagnosing these maladies required a postmortem brain biopsy—obviously, too late for patients. During postdoctoral work at Caltech in 1996,Lee identified a marker protein for sporadic Creutzfeldt-Jakob disease, yielding the first premortem test for the ailment. Lee went beyond the traditional method of studying a few proteins at a time; instead, he simultaneously analyzed the 2,000 proteins in human spinal fluid to pick out the telltale compound. In 1997 he confirmed that the disruptive protein also appears in mad-cow-afflicted cattle. People are now being tested for the protein in the U.S. and Europe. No one has confirmed whether the same marker characterizes variant Creutzfeldt-Jakob disease, but Lee’s team recently identified other protein indicators that may prove fruitful. Lee is also working on a similar test for Alzheimer’s.