Main Research Directions
Computational Biology. We are interested in protein structure, dynamics, expression, and function. To study these problems we usually integrate various forms of molecular and genetic data, carry out advanced analyses that help us understand phenomena of interest, and develop machine learning approaches to make predictions. Understanding protein function is a step towards understanding molecular mechanisms of human genetic disease.
Machine Learning. We are interested in supervised and semi-supervised learning and developing accurate models from positive vs. unlabeled data. Positive-unlabeled data comes in many forms and can pose additional questions such as how to address bias or how to learn structured outputs. We are also interested in the questions of accurate evaluation of machine learning methods, which is an under-appreciated aspect in machine learning.
Methods, Software, Code, Data
Over the years, we participated in the development of many methods, tools, databases, etc. You can access them from the Publications page (see Web, Code, etc. next to the publications).
Group Meetings, Journal Club
Spring 2017: Wednesdays 11-12:30 in Lindley Hall 101
Last modified: March 19, 2017