This application for the K99/R00 "Pathway to Independence" program aims to support Dr. Aurel Cami's development into an independent investigator in the field of Biomedical Informatics. The proposed career development period consists of a two-year mentored phase and a three-year independent phase. During the first phase, mentored by Drs. Reis, Kohane, Szolovits, and Manzi, the applicant will obtain formal training in drug safety and in biostatistics/machine learning methods for developing advanced predictive network models. This training will provide the groundwork for the proposed research project. The overall goal of this project is to develop novel surveillance approaches that leverage the inherent network structure of drug safety relationships across the entire drugome in order to predict adverse events earlier and more accurately. The project has four specific aims: 1. Construct and characterize integrated Drug Safety Network representations of the drugome. 2. Develop predictive network models to identify unknown drug-AE associations. 3. Develop predictive network models to identify unknown drug-drug-AE interactions. 4. Systematically evaluate model performance, both retrospectively and prospectively. The preliminary research studies have shown that predictive network models can sensitively and specifically detect drug-AE associations and drug-drug interactions, providing strong motivation for the proposed research. With a background in predictive network modeling and public health drug surveillance, Dr. Cami is highly qualified to conduct the proposed research. This project brings together a broad, interdisciplinary team of mentor, co-mentors and advisors with extensive combined experience in every aspect of the proposed research.
Some drugs that are approved for sale to the public may have dangerous unknown side-effects. It is important to detect these unknown side effects as soon as possible in order to prevent serious illness or death. This project will help protect the public health by improving the ability to detect unknown dangerous drug side effects earlier and more reliably.
|Cami, Aurel; Reis, Ben Y (2014) Concordance and predictive value of two adverse drug event data sets. BMC Med Inform Decis Mak 14:74|