The HuBMAP FUSION Research Experience was launched on March 17, 2025, as a remote, two-week asynchronous initiative designed to provide students with a unique opportunity to explore the cutting-edge intersection of artificial intelligence, bioinformatics, and spatial-omics data analysis. While the program was primarily designed for undergraduates, postdoctoral scholars, faculty, and other researchers were also welcome to apply and participate.
This flexible learning format allowed participants to engage with structured instructional modules and comprehensive guidelines at their own pace, while participating in interactive and collaborative research experiences. The course introduced students to FUSION’s functionality for visualizing and analyzing spatial transcriptomics data and developed their skills in using the “BulkLabels” component for annotating Functional Tissue Units (FTUs) based on predefined molecular or histological criteria. It also enhanced familiarity with HuBMAP’s Workspaces feature and its applications in bioinformatics research, provided hands-on experience in managing data visualization, and fostered interdisciplinary collaboration and scientific communication within research teams.
One of the central tools introduced during the program was FUSION, an interactive AI-powered interface that integrates spatial transcriptomics data with histopathology. FUSION enables segmentation of FTUs and links them with gene expression data, offering insights into both healthy and injured cell types, mapped directly onto histological features in Whole Slide Images (WSIs). These capabilities make FUSION a transformative tool for understanding tissue health and pathology.
Through active participation in HuBMAP projects and engagement with FUSION’s interactive tools and assignments, students gained valuable hands-on training, contributed to the development of cutting-edge research tools, and enhanced their skills in Python scripting, data visualization, and collaborative scientific inquiry. This program is ideal for learners seeking to bridge computational methods with biomedical research while supporting structured yet flexible learning. Additionally, 14 usability assessments were completed, providing valuable feedback to guide future improvements.
Stakeholders and Collaborators
UNIVERSITY OF FLORIDA
Pinaki Sarder
UNIVERSITY OF flORIDA
Yulia A. L. Strekalova
University of florida
Mishal Khan
