MIT EECS - Draper Laboratory Undergraduate Research and Innovation Scholar
Identification of genome-wide combinatorial regulatory structure using approximate inference
David K. Gifford
Recent cross-cell line data collection projects such as ENCODE has resulted in the discovery of thousands of putative functional elements in the genome. However, these annotations fail to identify the exact sequences that comprise a functional element, as well as their interactions. The goal of this project is twofold. First, to understand the minimal set of DNA sequences that control phenotypes observed by sequencing data, and second, to construct and understand the interaction between these sequence elements. Understanding these basic DNA elements and their interactions will allow us to understand simple transcriptional regulation at a genome-wide level.
My project lies at the intersection of biology, machine learning, and performance engineering. I worked on improving Google Searchs question answering system. I optimized and added new features to Googles next generation video codec. I have conducted research in manifolds, and taken several biology classes.