Building end-to-end machine learning systems for providing personalized health and ancestry insights
| Event Date: | April 6, 2023 |
|---|---|
| Time: | 7:30 pm |
| Location: | HAMP 2118 |
| Priority: | No |
| School or Program: | College of Engineering |
| College Calendar: | Show |
Discussion with 23AndMe Director Subarna Sinha
Host: Dr. Somali Chaterji, assistant professor of agricultural and biological engineering and electrical and computer engineering (by courtesy), CEO of KeyByte
Speaker via Zoom: Dr. Subarna Sinha, director of data and machine learning, 23andMe
Hosted by Innovatory for Cells and Neural Machines (ICAN) with the IEEE EMBS Midwest Region and Purdue Chapter, and the Purdue Society of Women Engineers
Tremendous advances in genomics over the last couple of decades have made it relatively cheap to identify DNA variants at millions of locations on the genome. 23andMe has accumulated large amounts of genetic data from their 13M+ customers, >80% of whom have consented that their data be used for research purposes. Besides genetic data, they also have non-genetic data that capture diverse information related to diet, lifestyle, disease status, etc. These data offer the unique ability to build machine learning models that can provide customers personalized insights about their health and/or ancestry. In this talk, we will provide a broad overview of the types of health and ancestry insights that can be generated using the data. We will also describe an end-to-end machine learning system that can train and deploy these models at scale.
