The U.S. National Institutes of Health (NIH)’s Office of Portfolio Analysis (OPA) recently announced the development of a machine learning model able to predict the likelihood of whether a scientific advance would make it to the clinic.
Recently published in PLOS Biology, Benjamin Ross, Editor with AI Trends recently reported on this important breakthrough.
Who Developed the AI Tool?
OPA Director George Santangelo and team
What is the Novel Metric?
The model qualifies predictions based on a new metric called “Approximate Potential to Translate” (APT).
As quoted by the authors, “We found that distinct knowledge flow trajectories are linked to papers that either succeed or fail to influence clinical research.” Hence, they concluded, “Translational progress in biomedicine can therefore be assessed and predicted…based on information conveyed by the scientific community’s early reaction to a paper.”
iCite V 2.0
NIH is launching version 2.0 of iCite, a web-based application offering a panel of bibliometric information for journal publications within a defined analysis group. The APT values, a product of the recent machine learning breakthrough mentioned above, will be freely available as a new iCite component.
Also available in V 2.0 is NIH’s Open Citation Collection (NIH-OCC) a free public access database for biomedical research. The database includes 420 million citation links with new citations added each month.
What are the benefits of Machine Learning in the clinical research process?
Well, as AI Trends notes, clinical research is a lengthy and cumbersome affair taking up to decades for discovery to “translate” to clinical studies. Santangelo noted to AI Trends that machine learning represented a significant opportunity for “a better read on the likelihood that papers would move into the clinic.
George Santangelo, Director OPA
Call to Action: To learn more, visit Office of Portfolio Analysis (OPA).