Tumor portraiture

In a pair of September papers published in Cell and Cancer Cell, researchers led by Ludwig Stanford’s Aaron Newman and Ash Alizadeh described their implementation of EcoTyper—a new machine learning framework that combines multiple algorithms previously developed by the researchers for the large scale analysis of cell types, genomic expression patterns in single cells and databases of cellular and molecular information—to categorize clinically relevant cell states and ecosystems of cells from tumor specimens at unprecedented scale. Applied to human carcinoma and diffuse large B cell lymphoma (DLBCL), the most common types of solid cancer and blood cancer, respectively, EcoTyper revealed a surprising diversity of cellular ecosystems in which distinct cell states interact with each other, including multiple ecosystems associated with different molecular subtypes and survival outcomes for patients. For example, EcoTyper defined new cell states and ecosystems in carcinoma, including ones associated with immunotherapy response and early lung cancer development. In DLBCL, it also identified five states of malignant B cells that vary in prognostic association and differentiation status, and significant variations in cell states for a dozen other lineages in the tumor microenvironment. The results paint a granular yet sweeping portrait of the microenvironment of human tumors and offer clues to new approaches to treating cancer.

This article appeared in the February 2022 issue of Ludwig Link. Click here to download a PDF (1 MB).


You are now leaving Ludwig Cancer Research's website and are going to a website that is not operated by the association. We are not responsible for the content or availability of linked sites. Do you wish to continue?