1 of 1

Vasculature Common Coordinate Framework Distance Visualizations Across Organs and Imaging Technologies

Data Generation and Interpretation

Colon, Tonsil, BE: Emma Marie Monte², Chenchen Zhu², John Hickey², Yuqi Tan², Bei Wei², Bingqing Zhao², Joanna Yang Bi², Garry P. Nolan², Michael Snyder² Lung: Jeffrey Purkerson³, Ravi Misra³, Gloria Pryhuber³ Spleen: Rafael dos Santos Peixoto⁴, Brendan F. Miller⁴, Jean Fan⁴, Maigan A. Brusko⁵, Todd M. Brusko⁵, Mark A. Atkinson⁵, Clive H. Wasserfall⁵ Skin: Fiona Ginty⁶ Colon: Clarence Yapp⁷, Jia-Ren Lin⁷, Peter Sorger⁷

​

Abstract

Considerations for using the vasculature as a cell coordinate system

Spatially registering human tissue samples in the HuBMAP EUI

A

C

E

Lung Cancer

Colon Cancer

Breast Cancer

Future Directions, References, and Acknowledgements

Expanding VCCF visualizations to other tissue types and imaging technologies

Figure 2. Vasculature Common Coordinate Framework, from [2].

The vasculature forms an uninterrupted path across scales in the human body, making it an ideal choice for creating a Common Coordinate Framework of the human body. The resulting Vasculature Common Coordinate Framework (VCCF) can localize cells of different types by using the nearest blood vessel that supplies it with oxygen. As part of the Human BioMolecular Atlas Program (HuBMAP), several tools have been built for spatially registering tissue samples and connecting them with expert ontologies via ASCT+B Tables in the Human Reference Atlas (HRA) framework. Interactive data visualizations show the distributions of distances between different cell types and their closest vasculature across organs and using different technologies. Here, we present Vitessce-based visualizations of 6 organs (skin, colon, Barrett’s esophagus, tonsil, spleen, and lung) and four technologies (CODEX, Cell DIVE, CyCIF, Xenium) from five different data providers. All datasets were RUI-registered (or are in the process of) and can be explored within the context of the 3D human body in the Exploration User Interface. We conclude with a discussion of planned extensions of the analysis and visualization workflows to cover disease (e.g., tumor cells) and hierarchical cell neighborhoods.

The spatial size, location, and rotation of tissue specimen are manually registered using the Registration User Interface (https://humanatlas.io/registration-user-interface) in coordination with data providers. All RUI-registered tissue blocks can be explored in the Exploration User Interface (EUI, https://apps.humanatlas.io/eui, Fig. 1) [1].

Figure 1. Exploration User Interface screenshots showing skin (Fig. 3) and colon plus spleen tissue registrations (Fig. 4).

1. Börner, K., Bueckle, A., Herr, B.W. et al. Tissue registration and exploration user interfaces in support of a human reference atlas. Commun Biol 5, 1369 (2022). https://doi.org/10.1038/s42003-022-03644-x
2. Weber, G. M., Ju, Y., & Börner, K. (2020). Considerations for Using the Vasculature as a Coordinate System to Map All the Cells in the Human Body. Frontiers in Cardiovascular Medicine, 7, 29.
3. Ghose, S., Ju, Y., McDonough, E. et al. 3D reconstruction of skin and spatial mapping of immune cell density, vascular distance and effects of sun exposure and aging. Commun Biol 6, 718 (2023). https://doi.org/10.1038/s42003-023-04991-z
4. Keller, Mark S., et al. "Vitessce: a framework for integrative visualization of multi-modal and spatially-resolved single-cell data." (2021).
5. Manz, T., Gold, I., Patterson, N.H. et al. Viv: multiscale visualization of high-resolution multiplexed bioimaging data on the web. Nat Methods 19, 515–516 (2022). https://doi.org/10.1038/s41592-022-01482-7
6. Hickey, J.W., Becker, W.R., Nevins, S.A. et al. Organization of the human intestine at single-cell resolution. Nature 619, 572–584 (2023). https://doi.org/10.1038/s41586-023-05915-x
7. dos Santos Peixoto, Rafael, et al. "Characterizing cell-type spatial relationships across length scales in spatially resolved omics data." bioRxiv (2023): 2023-10.
8. Currlin, Seth, et al. "Immune, endothelial and neuronal network map in human lymph node and spleen.�" bioRxiv (2021): 2021-10.
9. Brbić, M., Cao, K., Hickey, J.W. et al. Annotation of spatially resolved single-cell data with STELLAR.

Nat Methods 19, 1411–1418 (2022). https://doi.org/10.1038/s41592-022-01651-8

1. Lin, Jia-Ren, et al. "Multiplexed 3D atlas of state transitions and immune interaction in colorectal �cancer." Cell 186.2 (2023): 363-381.

References

Table 1. Required data format example.

Data Analysis and Visualization

Yashvardhan Jain¹, Yingnan Ju¹, Dan Qaurooni¹,

Andreas Bueckle¹, Katy Börner¹

¹Indiana University Bloomington ²Stanford University ³University of Rochester Medical Center ⁴Johns Hopkins University ⁵University of Florida ⁶GE Research Center ⁷Harvard Medical School

A Vasculature Common Coordinate Framework has been proposed to map all 37 trillion cells in the human body in a way that addresses its multiscale nature [2]. The vasculature seamlessly connects the macro-, meso-, and micro-scales of the body and hence provides an ideal pathway to assign “zip codes” to these cells in order to localize them, see Fig. 2. Ghose S., Ju Y., et al. [3] looked at the distance distributions between different immune cell types and the closest endothelial cells in 3D reconstructed tissue samples from adult human skin tissue using Cell DIVE, see Fig. 3. This enabled an in-depth analysis of different distance distributions focussed on the effects of UV sun exposure and aging in three dimensions.

Figure 3. 3D VCCF Visualizations of skin Cell DIVE data, from [3].

The visualization workflow has been generalized to cover more tissue types and assay type technologies from different data providers. Furthermore, the open-source visualization tool Vitessce [4,5] can now be used to explore 2D VCCF visualizations within the HRA Portal. Fig. 4 shows first results on colon (CODEX [6]) and early stage colon polyp (Xenium; HTAN) datasets from Stanford University, spleen (CODEX [7,8]) datasets from University of Florida and Johns Hopkins University, tonsil (CODEX [9]) and Barrett's esophagus (BE) (CODEX [9]) from Stanford University, Lung (CODEX) from University of Rochester Medical Center, and colon (CyCIF [10]) from Harvard Medical School (HTAN).

2000 μm

Figure 4. VCCF Visualizations of 8 datasets across organs and imaging technologies.

In close collaboration with different HuBMAP and HTAN tissue data providers, �we will enhance the visualization workflows based on user needs, e.g., to support more in-depth analyses of the vascular system in correlation to different cell types across organs; making it possible to pick “anchor” cell types to visualize distance to cell types other than endothelial cells; adding scale bars, legends, and distance distribution histograms within the Vitessce viewer; visualizing 3D data in Vitessce as shown in Fig. 3 and analyzing/seeing all distances for a selected cell, cell type, or cell neighborhood (e.g., FTU); to add imaging data and turn specific image channels on/off; and compute quantifications of cell-type colocalization as a function of the z-plane in 3D datasets.

Barrett’s Esophagus - CODEX - 7µm thickness - 0.377µm per pixel

Tonsil - CODEX - 7µm thickness - 0.377µm per pixel

Colon - Xenium - 5µm thickness - 0.2125µm per pixel

Spleen - CODEX - 5µm thickness - 0.377µm per pixel

Colon - CyCIF - 5µm thickness - 0.65µm per pixel

Lung - CODEX - 5µm thickness - 0.5056µm per pixel

Lung - CODEX - 5µm thickness - 0.5056µm per pixel

Skin - Cell DIVE - Male/Female

Colon - CODEX - Male/Female

Spleen - CODEX - Male/Female

Colon - CyCIF - Male

500 μm

250 μm

250 μm

250 μm

50 μm

50 μm

250 μm

2000 μm

2000 μm

500 μm

500 μm

250 μm

500 μm