Mission

Improve human health

The core mission of the Division is to alleviate human suffering by reducing the burden of diseases on individuals and on the population. This mission informs all our activities in developing and applying computational technologies. Our technological capabilities allow us to impact a broad range of human diseases including infectious, cancer, heart, kidney, intestinal, autoimmune, allergies, and neurological disorders.

Advance the field of pathology

Pathology is both a scientific and a medical discipline. As a scientific discipline, it involves the study of diseases, including their basic mechanistic causes and effects. As a medical discipline, it focuses on the diagnosis of disease through tissue and fluid samples. Our goal is to advance the field of pathology through computational technologies, to improve the understanding, diagnosis, and treatment of human diseases. With this broad view of pathology, we work on a variety of applications such as deep learning/artificial intelligence to improve cancer diagnosis/prognosis from histology images and to create new live bacterial therapeutics to treat infectious or autoimmune diseases.

Develop innovative computational methods

Human diseases often have complex causes and effects on the body. Data needed to analyze human diseases is similarly complex and multi-faceted, and may include imaging, nucleic acid sequencing, mass spectrometry, and many other modalities. Unlike in classic engineering applications, these data are often challenging to acquire (e.g., due to costs, rarity of the disease, etc.), and thus dataset sizes are by nature often limited. These and other challenges necessitate going beyond application of existing computational methods. Thus, we actively engage in computational research, to develop novel computational models, inference algorithms and integrated pipelines. To accomplish this, we leverage a variety of advanced computational disciplines, including Bayesian nonparametric statistics, deep learning and control theory.