Project Spotlight: CADidME

Vanathi Gopalakrishnan, PhD, is the Director of the PRoBE laboratory for Pattern Recognition from Biomedical Evidence within the University of Pittsburgh’s Department of Biomedical Informatics. She works with the lab’s teams to design and develop novel machine learning algorithms using symbolic, probabilistic, and hybrid approaches to solve bioinformatics problems including biomarker discovery and disease classification. Dr. Gopalakrishnan is also a lead researcher on a Center for Commercial Applications of Healthcare Data funded project: CADidME.

Where did the CADidME project begin?

The CADidME project began with a fabulous collaboration between me and the founding director of Pitt’s Clinical and Translational Science Institute (CTSI), and 2018 Health Care Hero Awardee, Dr. Steven E. Reis. Steve has been collecting vast amounts of research data over the past fifteen years in a prospective study – Heart SCORE – designed to identify and reduce racial disparities in cardiovascular disease (CVD). The Heart SCORE study has amassed a wide array of phenotype, imaging, participant provided information, genotype, and metabolomics data as well as clinical events from 2,000 participants who have been followed annually since 2003. Steve and I had discussed prior to the generation of the metabolomics data, that the PRoBE lab would help with analysis of the data. That was when the idea for the CADidME project started to develop in my mind. We applied for Pittsburgh Health Data Alliance funding and were very fortunate to be selected. The funding was absolutely crucial for my laboratory as it enabled us to assemble a highly skilled team of researchers and developers to conduct analyses on these data for predicting Major Adverse Cardiovascular Events (MACE).


Walk us through your project.

CADidME stands for “Coronary Artery Disease intelligent detection via Metabolomic Expression”. The long-term goal of this project is to be able to develop personalized risk scores for cardiovascular disease (CVD) adverse events in order to enable the precise management of individual patients. My laboratory specializes in applying computational techniques to data to learn classification rules. In this current phase of CADidME we have used the metabolomics and other data to develop multiple risk prediction models that are tailored to deliver predictions relevant to patient subgroups with known, underlying risk factors.


What are your project’s next steps?

In Phase II of this project, we are positioning ourselves to develop the personalized risk score based on the analyses from the previous phase. Our team has also developed novel software programs to help us validate the discoveries that have arisen from CADidME and we are working on future directions that could be used to discover new treatments or therapeutics. The PHDA will be a wonderful partner to assist our team towards these next steps leading to commercialization, as we have formed a great working relationship with every member of this extended team, who provides our project with significant feedback to continuously focus our efforts.


What do you foresee the future of innovation looking like here in Pittsburgh?

Pittsburgh is proving to be the next epicenter of technology. Especially in the biomedicine and health space, we’re leading the way and the Alliance is playing a key role. Over the next several years, I see the PHDA as an organization that will empower innovative technology startups and continue to: (1) provide clinicians with better tools for collaborative care of patients; (2) impact how we measure, track, and transform lifestyle-related behavior to advance individual health; (3) investigate how new therapies can be used to prevent health issues as part of a toolkit of cultural practices promoting well-being; and (4) make discoveries related to immunotherapies for advanced cancer care.