Systems Biology Group Meeting

Presenter: Meet Barot, New York University

Title: Unsupervised Classification of Proteins

Abstract: With the large amount of sequence information constantly being generated, it has become an important problem to automatically cluster proteins into meaningful categories without label information. An advancement on this task could enable discovery of new functional categories from sequence and network information alone.

In this talk, I will present my ongoing work on clustering protein sequences without label information into protein families. I will also show some results for clustering proteins into enzyme classes from the BRENDA enzyme database using both sequence and protein-protein interaction network information.