Will data democratization truly transform healthcare? According to a representative from a future generation of healthcare disruptors, the answer is clearly yes. In 2014, high school student and HAS 19 keynote presenter Justin Aronson launched Variant Explorer, a cloud-based data integration and visualization system for improving the clinical interpretation of sequenced genetic variants. Aronson credits expert mentors (geneticists Heidi Rheim and Steven Harris) with his achievement, but he says that the real determinant in Variant Explorer’s success is the data he accessed thanks to his data resource’s commitment to open sourcing its asset.
Aronson built Variant Explorer on genomic and phenotype data from the archive ClinVar. As a freely accessible, public record of reports of the relationships among human variations and phenotypes, ClinVar provides public access to information about the relationships between human variation and observed health status. Through displays and data processing, Variant Explorer helps healthcare organizations identify variations in relationships (i.e., conflicts between labs) among the genomic variants and phenotypes, which labs, researchers, and other clinicians submit to ClinVar.
The Variant Explorer homepage prioritizes user experience with three accessible tabs tailored to users (Search by Submitter, Show Submitter Mega Table, and Search by Significance). The tabs make data accessible, and a conflict discrepancy table helps break down data and shows conflict with a high-level overview. Labs visualize and recognize conflicts with other labs and can fix them within own organizations or reach out to others to fix conflicts, with the end goal of improving patient care.
Aronson is now looking to machine learning to increase Variant Explorer functionality. Specifically, he aims to reformat ClinVar data to make it more conducive to machine learning and then leverage the advanced analytics capability to show more sophisticated relationships between disparate lab data. Aronson says that ClinVare data is currently extremely complex, and while a great resource for labs to see a lot of information, it doesn’t support machine learning.
By simplifying ClinVar data structure for machine learning use with Variant Explorer, Aronson would also open source the data for other users to leverage machine learning. Skeptics in the data community have told Aronson that his machine learning goal is too ambitious, but he’s moving forward with an eye to the future. “Maybe it doesn’t work out,” he says, “but I think that even by failing at incorporating machine learning in Variant Explorer, I’ll learn a lot more about machine learning and be able to use it for [other] projects.”
Aronson explains that the ClinVar commitment to data democratization has not only enabled his work with Variant Explorer but serves as a great example of how open-source data is key to solving and facing the increasingly complex problems of today and future. “I hope these larger corporations embrace data democratization because I think it’s key to how we, as a society, will move forward,” Aronson says.
Would you like to learn more about this topic? Here are some articles we suggest:
Would you like to use or share these concepts? Download the presentation highlighting the key main points.