BEHAV3D Tumor Profiler (BEHAV3D-TP) is a cloud-based computational platform designed to comprehensively map and analyze the dynamic heterogeneity of tumor cells within their tumor microenvironment (TME). Built as a modular Jupyter Notebook with an intuitive graphical user interface, BEHAV3D-TP runs directly in a web browser via Google Colaboratory, making it freely accessible and removing the need for local software installation. This design ensures that researchers, regardless of programming expertise, can analyze single-cell dynamics from live imaging data.
BEHAV3D TP allows for unbiased classification of individual cells from intravital microscopy time-lapse data based on a combination of morphological, environmental, and dynamic features. The pipeline begins with segmented and tracked time-lapse image data and applies the heterogeneity module, the core of BEHAV3D-TP, which uncovers diverse morphodynamic behaviors of tumor cells. Two extension modules expand this functionality: the large-scale phenotyping module, which integrates tumor behavioral patterns with tissue phenotyping from post-IVM fixed imaging to study behavioral distributions across TME regions; and the small-scale phenotyping module, which uses in situ-labeled TME structures from IVM data to link local tumor cell behaviors with their immediate microenvironment. Together, these modules provide a multi-scale framework to dissect how different TME features influence tumor cell dynamics, offering broad accessibility and powerful insights into tumor–TME interactions.
