As one of the largest centers dedicated to providing biomedical research and training at NIH (National Institutes of Health), NIGMS (National Institute of General Medical Sciences) supports over 3,300 principal investigators per year on R01 research project grants. In order to improve stability of research funding for principle investigators, deliver more efficient budget planning and recognize funding opportunities in the future, NIGMS launched a pilot program to further understand the pool of grantees who was in short funding gaps.
This project identifies five key factors that have the most significant effect on principal investigators’ re-entry or exit behavior after gaps and implements a Multivariate Logistic Regression model for predicting the probability of a principal investigator dropping out of the funding pool in different time frames given a specific date proposed. These findings will help NIGMS estimate the grant application volume, detect the principal investigators who might be at a higher risk of losing funds, and explore efficient ways to allocate biomedical funds for long- term planning purposes. Still, there are constraints need to be overcome for better application and prediction.