We’re excited to share our lab’s first manuscript, now available on bioRxiv! In our study, Tumor Landscape Analysis: An Ecologically Informed Framework to Understand Tumor Microenvironments, we introduce the Tumor Landscape Analysis (TLA) pipeline—a novel computational approach designed to rethink how we study cancer tissues.
Tumors aren’t random collections of cells; they’re dynamic, evolving ecosystems. Inspired by principles from landscape ecology, TLA allows us to quantitatively capture the spatial organization of tumor microenvironments (TMEs) with a level of detail and interpretability that has been hard to achieve with traditional methods. Rather than treating cells in isolation, TLA looks at the broader architecture, the “ecological landscape”, of tumors, offering methods to measure cellular distributions, tissue fragmentation, and microenvironmental niches.
At the heart of TLA is a powerful set of metrics adapted from ecology. By applying tools like the Morisita-Horn index, Ripley’s H function, and Shannon diversity, we can quantify how tightly different types of cells cluster together, how evenly they are distributed, and how organized—or disorganized—the tissue has become. Importantly, TLA is imaging-agnostic and flexible: it works with whole-cell, point-based, or region-level data across different tumor types and sample formats.
One unique feature of TLA is the identification of local microenvironments (LMEs). Instead of relying on pre-defined biological categories, TLA uses data-driven spatial patterns to map tumor landscapes into reproducible niches, based on local cell type abundance and mixing behavior. These ecological “zones” can help us understand, in a completely unsupervised way, how tumors evolve, adapt, and respond to therapy.
Why does this matter? Because spatial features of the TME—such as cellular diversity, spatial clustering, and landscape fragmentation—are increasingly recognized as critical drivers of therapeutic resistance and disease progression. By translating the physical layout of cells into quantitative, analyzable data, TLA provides a new lens for exploring tumor biology, stratifying risk, and potentially guiding precision oncology strategies.
This work reflects a true team effort, combining expertise in oncology, spatial statistics, computational biology, and evolutionary theory. We are grateful for the support from the Arizona Cancer Evolution Center, Gerstner Family Foundation, the Grand Forks Career Development Award, and Mayo Clinic’s Center for Clinical and Translational Science (CCaTS).
If you’re interested in applying TLA to your own spatial datasets—or just want to learn more—the full pipeline is freely available on our GitHub. We’re looking forward to collaborating with others who share our passion for decoding the spatial complexity of cancer!
Stay tuned as we continue to expand this framework and explore its applications across different cancer types and treatment contexts.
Read the full preprint: Tumor Landscape Analysis on bioRxiv
Leave a Reply