James Ko Email & Phone Number

New York, New York, United States

Build and demonstrate ETL pipeline from Cancer Research Data Commons (CRDC) to cBioPortal cancer visualization platform under joint collaboration with National Cancer Institute (NCI), Institute for Systems Biology (ISB), and Leidos, making data accessible to thousands of researchers. Extract clinical and genomics data from CRDC via Google BigQuery database and convert to format usable by cBioPortal. Help integrate data pipeline with Apache Airflow to achieve robust error reporting and task parallelization. Work with backend technologies including Python, SQL, Google BigQuery, Apache Airflow, and Pandas.

Boston, Massachusetts, United States

Managed data for DISCOVERY and MarkVCID studies on cognitive impairment and dementia, with goal of developing effective biomarkers and personalized medicine for patients. Created weekly reports summarizing DISCOVERY site enrollment statistics and upload rates for researchers, helping to identify sites lagging in data uploads. Collected and anonymized clinical form data for MarkVCID1 into spreadsheets, enabling researchers outside consortium to conduct data analyses. Gained experience with backend technologies such as Python, CouchDB, Bash, and Cron.

Boston, Massachusetts, United States

Worked with Harvard Medical School doctor to build PyTorch machine learning model to predict biomarkers of neurological conditions such as Alzheimer's disease and schizophrenia. Helped other interns integrate model into projects studying effects of COVID-19 on brain aging and congenital heart disease. Wrote scientific paper analyzing effect of data imbalance and sample size on model performance and submitted for review to NeuroImage.