CMI Lab Research Featured in Nature Collection

We are thrilled to announce that the research conducted by the CMIL has been showcased in the esteemed Nature collection. The spotlight article, titled “A user-friendly tool for cloud-based whole slide image segmentation with examples from renal histopathology,” sheds light on HistoCloud, our very own open-source cloud tool for pathology image segmentation. Our findings show that HistoCloud is accurate, easy to use, and comparable to human expertise. This fair-use software is accessible via the Athena server hosted on UF’s HiPerGator supercomputer. While the tool is currently in its alpha version, the CMI Lab will soon release it into production and offer dedicated technical support to the entire community.

The HistoCloud article is amongst a set of nine scientific papers released in the Nature family of journals and the journal Cell Reports describing breakthroughs in high-resolution, multiscale, and multimodal mapping of the human body in two and three dimensions. These new articles are part of a larger tranche of papers that present the first collection of maps generated by scientists at institutions supported by the Human BioMolecular Atlas Program (HuBMAP), a consortium managed by the NIH Common Fund.

HuBMAP operates through a collaboration between the National Institutes of Health (NIH) and HuBMAP Integration, Visualization, and Engagement (HIVE) teams, Tissue Mapping Centers (TMCs), Tissue Technology Development (TTD) teams, Rapid Technology Implementation (RTI) teams, and Demonstration Project (DP) teams. The CMI Lab serves as one of the HIVE teams focused on developing artificial intelligence and machine learning tools to map spatial omics data on brightfield histology. Although not featured in the Nature package, the CMI Lab has published other articles on works conducted as part of the HuBMAP consortium. The article, PodoCount: A Robust, Fully Automated, Whole-Slide Podocyte Quantification Tool, pertains to a machine learning tool we developed to estimate podocytes at the single cell level from brightfield immunohistochemistry whole slide images.

HuBMAP is developing an open framework to map the molecular landscape of healthy cells, as well as their immediate surroundings, in the human body to better understand the relationship between cellular organization and health. HuBMAP researchers use the latest single-cell imaging methods to characterize major structures and organs in the human body. Atlases of healthy tissues play a vital role in biomedical research by serving as a reference that can be compared to diseased samples. These atlases may help scientists to understand what goes wrong earlier in disease progression. This knowledge can provide key insights into developing treatments early in a disease.  Healthy tissue is rarely studied, making the work of HuBMAP absolutely critical.

“This is the first coordinated set of papers to chart out a set of high-resolution, single-cell maps of human organs, including the functions being carried out,” said Michael Snyder, Ph.D., principal investigator of one of the HuBMAP mapping projects. “Much as building a working engine requires you first to assemble parts into working components and not just throw individual parts together, our investigators’ work has uncovered how single cells in the body live and work together in what we call ‘cellular neighborhoods.’ The insights provided by these discoveries represent the first step in creating the framework for a meaningful three-dimensional, functional atlas of cells in the human body.”

The paper package is complemented by a “Perspectives” article in the journal Nature Cell Biology that delineates the overall progress made so far by HuBMAP since it started in 2018, as well as the program’s goals for the next four years.

This collection of papers reports the latest progress by the HuBMAP consortium, which consists of more than 400 scientists at more than 40 institutions in the U.S. and across the world. These investigators bring valuable and varied expertise to produce results that could not have been accomplished by a single lab. A more complete list of consortium publications can be found here.

“NIH is investing $215M over 8 years through the HuBMAP Consortium to build this strong framework of cutting-edge, high-throughput sequencing, and imaging techniques to understand the role cellular organization plays in human health,” said Dr. Richard Conroy, the HuBMAP Program Leader, at the NIH. “We are delighted by this first set of maps and we look forward to the consortium expanding and deepening the range of tissues maps available that we hope will enable new discoveries over the next 4 years.”

HuBMAP’s successes in the last five years include:

  • Mapping the cellular organization of the human intestine, kidney, liver, the interface between fetus and mother, and skin at the single-cell level
  • Identifying cell types and extracellular material associated with functional tissue units or cellular neighborhoods
  • Identifying new cell types or previously unrecognized functions of cells 
  • Establishing an infrastructure to support mapping the human body at single-cell resolution. The framework will include data standards, reference materials to help classify cells, and the molecular probes used to study them 
  • Creating and validating a suite of tools, to  improve the identification and labeling  of cells
  • Building tools such as Azimuth, Human Reference Atlas Portal, and Vitessce that allow users to analyze, visualize, and compare their data to these references
  • Establishing cross-consortia collaborations to align atlas outputs to interpret health and disease
  • Strategic partnerships with industry leaders to create reporting standards for mapping technologies

In the future, HuBMAP researchers will continue to collect, analyze, and visualize cellular organization data for other healthy human organs, such as bladder, eye, female reproductive organs, heart, knee,  lung, lymph nodes, pancreas, spleen, and thymus.  They will also continue to build the tools and technologies necessary to construct the framework needed for a reference atlas of the human body at cellular resolution. 

About HuBMAP: The Human BioMolecular Atlas Program (https://hubmapconsortium.org) is an NIH Common Fund program working to catalyze the development of a framework for mapping the human body at single-cell resolution. HuBMAP values secure, open sharing, and collaboration with other consortia and the wider research community. HuBMAP’s online Data Portal offers access to the consortium’s data, as well as to the tools developed by its investigators.

Research reported in this publication was supported by the Office of The Director, National Institutes Of Health under Award Numbers OD033753, OD033756, OD033758, OD033759, OD033760, OD033761, HL145593, HL145600, HL145609, HL145623, CA255132, CA255133, CA255134, CA255135, AI142766, DK120058, HG010426, HL145608, HL145611, OD026663, OD026671, OD026673, OD026675, OD026677, OD026682, CA246594, CA246632, CA246633, CA246635, CA256959, CA256960, CA256962, CA256967, CA257393, AR078664, DK127823, EY032442, HD104392, HD104393, HL156090, AR081774, AR081775, DK134301, DK134302, HD110347, HG012723, HL165440, HL165442, HL165443, HL165445, HD110336, HG012680, HL166058, 1UG3CA256962, 4UH3CA256962, and HL166060.