NaroNet: Discovery of tumor microenvironment elements from highly multiplexed images
This work addresses the problem of discovering unknown TMEs for cancer researchers, offering an accessible tool that accelerates biomarker signature discovery, though it appears incremental as it builds on existing computational methods for multiplex image analysis.
The paper tackles the challenge of identifying novel tumor microenvironment elements (TMEs) from multiplex immunostained tissues by presenting NaroNet, a machine learning method that discovers and annotates TMEs from self-supervised cell embeddings and uses their abundance for patient classification, validated on synthetic and real cancer datasets where it successfully identifies relevant TMEs.
Many efforts have been made to discover tumor-specific microenvironment elements (TMEs) from immunostained tissue sections. However, the identification of yet unknown but relevant TMEs from multiplex immunostained tissues remains a challenge, due to the number of markers involved (tens) and the complexity of their spatial interactions. We present NaroNet, which uses machine learning to identify and annotate known as well as novel TMEs from self-supervised embeddings of cells, organized at different levels (local cell phenotypes and cellular neighborhoods). Then it uses the abundance of TMEs to classify patients based on biological or clinical features. We validate NaroNet using synthetic patient cohorts with adjustable incidence of different TMEs and two cancer patient datasets. In both synthetic and real datasets, NaroNet unsupervisedly identifies novel TMEs, relevant for the user-defined classification task. As NaroNet requires only patient-level information, it renders state-of-the-art computational methods accessible to a broad audience, accelerating the discovery of biomarker signatures.