Deep Learning in Genomics


Can deep learning models that have defeated gamers or recognized images better than humans also help us understand genomics? How far will this interdisciplinary research take us on our quest to cure cancer? In an era with faster-than-Moore’s-Law exponential growth of the genomics data (Berger et al. 2016), deep learning methods are finally able to assist in solving essential problems in the field. However, these exciting developments also face challenges that are unique to working with data from our DNA.

As researchers trying to combine deep learning and genomics, we have to think carefully about applying these models effectively to genomics tasks. Is it appropriate to use deep learning for our application? What model should we use? Will our approach improve our understanding of the data or the problem? In this course, you will answer these questions by our coverage of recent research literature in the class. You will learn about different genomics tasks, deep learning models, and how they fit together. The course is designed to enable critical thinking and allows students to work together to apply these models.

Course announcements will be made on Piazza, assignment and lecture materials will be uploaded on the course website, and assignments will be submitted and graded on Gradescope.

Professor: Ritambhara Singh (ritambhara_singh@brown.edu)
Time & Location: TF 4:00pm – 5:20pm in CIT 101

Documents: Course Missive
Piazza: piazza.com/brown/spring2020/csci1850
Gradescope: https://www.gradescope.com/courses/82080/
Contact HTAs only: cs1850tas@lists.brown.edu
Contact HTAs + Professor: cs1850headtas@lists.brown.edu


The class goes virtual starting from March 13 (Friday). Please refer to the Piazza post for details.