For years, critics have argued that the U.S. National Institutes of Health favors clinical research over fundamental science. A new analysis of more than 2.2 million grant applications suggests they may be right—and indicates that the shift has been driven less by an increase in clinician-scientists pursuing research than by administrative decisions at the agency.

The comprehensive study, spanning fiscal years 1986 through 2017, shows that while support for curiosity-driven research through the traditional R01 grant remains prominent, NIH has increasingly channeled funding through alternative types of grant mechanisms. Although many of the new mechanisms theoretically support the entire spectrum of biomedical research, in practice, awarded applications have increasingly pursued translational and clinical science.

A team of researchers has developed a novel network-based strategy to predict which areas of biomedical research will produce transformative breakthroughs—up to twelve years before the key discoveries are published and 30 years or more in advance of their recognition by a major prize. The work demonstrates that scientific progress, long thought to be unpredictable, leaves measurable traces well before its impact becomes widely recognized.

The study, Prediction of transformative breakthroughs in biomedical research, analyzed the co-citation network of more than 17 million papers in PubMed, the comprehensive database of peer-reviewed biomedical literature. Using this network to map distinct scientific topics, the researchers isolated and tracked the work that produced past breakthroughs, discovering a common signal that preceded more than 90 percent of Nobel Prize– and Lasker Award–winning biomedical discoveries included in their analysis.