Headshot portrait of Maxim Zaslavsky - Bio-X Bowes Fellow
Bio-X Graduate Student Fellow

Awarded in 2022
Home Department: Computer Science
Faculty Advisors: Anshul Kundaje (Genetics and Computer Science), Scott Boyd (Pathology), Robert Tibshirani (Biomedical Data Science and Statistics), and Benjamin Pinsky (Pathology and Medicine - Infectious Diseases)

Research Title: Universal Disease Diagnostics with Immune Repertoire Sequencing and Machine Learning

Research Description: The Kundaje, Boyd, Tibshirani, and Pinsky labs and Maxim introduce a method to find the target pathogens that human immune receptors can pattern match, using deep learning language models of immune sequence patterns. In a proof of concept, their strategy distinguished five immune states spanning a wide range of diseases— from acute to chronic infections, to autoimmune and immunodeficiency disorders — with a high level of accuracy. This interpretable model of immune receptor sequences independently recapitulates known SARS-CoV-2 biology and enables tracking immune response changes during disease progression. The proposed research plan will develop this proof of concept of deciphering disease exposures into a pan-disease diagnostic, gleaning medically useful information from immune cell populations.