Launching patient-partnered research programs
After my time teaching, I turned down a full time classroom position and took a product ops/product manager role at the Broad Institute of MIT and Harvard to launch a nonprofit called Count Me In (CMI). Put simply, CMI is a clinical and genomic marketplace that facilitates data exchange between patients and researchers (akin to Uber’s marketplace that connects drivers and riders or Amazon’s marketplace connecting sellers and buyers). We had a patient and hospital facing product which represented the data supply and a public real-world clinical and genomic database which represented the data demand.
During my tenure at CMI, I worked on data supply and demand products. I started by building the data supply product alongside patients, software engineers, physicians, regulatory specialists and lab ops. I helped build a new enrollment website, medical record request pipeline and sample collection system that would work for any patient in the United States or Canada. I led user interviews with patients and physicians to design the MPCproject and ESCproject websites and incorporated key findings into our product (e.g. new password-protected enrollment workflow). I am co-first author on a paper in Cell Genomics describing the operations and early results of the MPCproject: A patient-driven clinicogenomic partnership for metastatic prostate cancer.
Mid-way through my time at CMI, I identified a need to have a technical product manager translating between software, lab ops and researchers to develop the data demand products, specifically the internal data platform and dataset generation pipeline. I took computational biology and programming classes at Harvard Extension School and taught myself how to code after work. I designed a sabbatical program within CMI to enable employees to develop needed skills to accelerate CMI’s mission and took my own sabbatical to embed with Eli Van Allen’s computational lab at Harvard Medical School. There I learned how to design and run variant calling pipelines for whole exome sequencing data. When I came back to CMI, I used these new skills to facilitate further automation of the data pipeline in partnership with our software engineering team. The new features I spearheaded included: 1) building a private instance of a www.cbioportal.org (a public data repository) to streamline QC of processed genomic data and 2) automate population of key hospitals from a CMS database to accelerate hospital data retrieval.
