Tumors contain many different types of cells organized in complex spatial patterns that can influence how the disease progresses. Because of this, it is hard to predict how a tumor will develop and respond to treatment. Researchers at the University of Illinois Urbana-Champaign are taking a new approach that combines geographic mapping techniques with gene expression analysis to visualize these spatial relationships inside tumors.
Cancer Center at Illinois (CCIL) Associate Director for Education Zeynep Madak-Erdogan collaborated with Illinois researchers, including some from the Carle Illinois College of Medicine, to better understand how tumor cells are organized. Their study, In Silico Reconstruction of Primary and Metastatic Tumor Architecture using Geographic Information System-Augmented Spatial Transcriptomics, introduces a computational framework called GIS-ROTA that helps map biological activity within tumors.
Cancer Center at Illinois Associate Director for Education Zeynep Madak-Erdogan
“To have such high-impact work, there should be research teams that include people from different sides of the sciences: cancer expertise, computational expertise, biology, and now AI and analytics tools.”
ZEYNEP MADAK-ERDOGAN
“We are studying tumors and, in this particular paper, breast cancer,” said CCIL member Aiman Soliman, Senior Research Scientist at the National Center for Supercomputing Applications. “Typically, you might expect spatial scientists to look at maps at the urban scale, but the question came up: why not take the same spatial analysis and apply it within the tumor?”
Jin Young Yoo, a graduate student in the Women’s Health and Metabolism Lab explained, “In this research, we developed an analytic framework for spatial-omics data that incorporates geospatial features to visualize and quantify the spatial relationships between different cell groups within tumors.”
Using this idea, the team developed Geographic Information System-Augmented Spatial Transcriptomics (GIS-ROTA). Instead of first grouping cells using statistical patterns and then trying to determine their biological meaning, the researchers started by looking at known biological pathways and asked where those functions appear within the tumor.
“Although all cells contain the same DNA, what defines tissue function is how genes are expressed and regulated,” said Yoo. “Our method maps the spatial activation of biological pathways, such as metabolism or immune response, rather than just grouping cells by similarity. This gives us clearer insight into tumor function.”
When the researchers applied their framework to estrogen receptor-positive breast cancer samples, they discovered significant spatial patterns.
“In this study, we looked at what changes are happening in the relative localization of the cells when we compare primary tumors with those that metastasize,” Madak-Erdogan said. “We identified different cell types and molecular pathways, which gives us a tool to target these pathways and hopefully make these tumors respond to therapies.”
By mapping where biological pathways are active within tumors, researchers hope to better understand mechanisms such as endocrine resistance that can limit treatment effectiveness in metastatic breast cancer.
“As clinicians, we see every day how endocrine resistance leaves patients with metastatic breast cancer with fewer treatment options,” said Dr. Maria Grosse Perdekamp, Clinical Assistant Professor of oncology at Carle Illinois College of Medicine. “Working with this research team allowed us to dig into the biology behind this problem in a completely new way. Our new approach gave us insights that can only come when the clinic and the lab work together.”
The team hopes this framework will become a useful tool for other researchers studying cancer. Because the method begins with known biological functions, it helps scientists more clearly interpret the development and function of the cells inside of tumors.
“This tool is flexible and biologically intuitive, and it’s not limited to one type of cancer,” Soliman said. “Researchers with different questions can use existing curated molecular libraries to ask where these circuits are mapped in their tissues, leading to new discoveries.”
Madak-Erdogan emphasized that projects like this require collaboration across many scientific disciplines.
“To have such high-impact work, there should be research teams that include people from different sides of the sciences: cancer expertise, computational expertise, biology, and now AI and analytics tools,” Madak-Erdogan said. “That’s quite important.”
Zeynep Madak-Erdogan
Sylvia D. Stroup Scholar and Professor of Food Science and Human Nutrition
CCIL Research Program and Theme
- Program: Cancer Engineering and Biological Systems
- Theme: Mechanistic and Quantitative Biology
Research Focus
Zeynep Madak-Erdogan improves the quality of life for postmenopausal women and breast cancer survivors by understanding how diet and nutrition affect hormone action. Her lab uses multiscale modeling of omics data from patient samples, animal models, and cell lines to understand the molecular basis of nuclear receptor-mediated metabolic regulation and of therapy resistance.
Aiman Soliman
Research Assistant Professor, National Center for Supercomputing Applications
CCIL Research Program and Theme
- Program: Cancer Technology and Data Science
- Theme: Computational Engineering and Data Science
Research Focus
Aiman Soliman uses big data to study the spatial organization of natural and urban environments. He also evaluates the reliability of different big data sources in analyzing spatial phenomena.
Editor’s notes:
Zeynep Madak-Erdogan is the Sylvia D. Stroup Scholar and Professor of Food Science and Human Nutrition at the University of Illinois Urbana-Champaign. She also serves as Associate Director for Education at the Cancer Center at Illinois is affiliated with Nutritional Sciences, Biomedical and Translational Sciences, the Carl R. Woese Institute for Genomic Biology, and the National Center for Supercomputing Applications.
She can be reached at zmadake2@illinois.edu.
Aiman Soliman is a Senior Research Scientist at the National Center for Supercomputing Applications and a Research Assistant Professor of Urban and Regional Planning at the University of Illinois Urbana-Champaign. He also serves as an assistant professor in Food Science & Human Nutrition department. His research focuses on spatial data science, computational methods, and spatial analysis.
He can be reached at asoliman@illinois.edu.
Jin Young Yoo is a Ph.D. student in Food Science and Human Nutrition in the Women’s Health and Metabolism Lab at the University of Illinois Urbana-Champaign and a T32 Tissue Microenvironment (TiME) trainee. Her research focuses on spatial transcriptomics and computational approaches to study tumor heterogeneity in metastatic breast cancer.
She can be reached at jyoo19@illinois.edu.
Maria T Grosse Perdekamp is a Clinical Assistant Professor of oncology at Care Illinois School of Medicine. She also serves as a clinician at Carle Cancer Institute, where she specializes in cancer, hematology, and oncology.
She can be reached at mtgp@illinois.edu
The paper, “In Silico Reconstruction of Primary and Metastatic Tumor Architecture using Geographic Information System-Augmented Spatial Transcriptomics,” published in Cancer Research, is available here.
DOI: https://doi.org/10.1158/0008-5472.CAN-25-3161
This story was written by Hailee Munno, CCIL Communications Intern.