ScopeViewer: A Browser-Based Solution for Visualizing Spatial Transcriptomics Data

  1. Danni Luo
  2. Sophie Robertson
  3. Yuanchun Zhan
  4. Ruichen Rong
  5. Shidan Wang
  6. Xi Jiang
  7. Sen Yang
  8. Suzette Palmer
  9. Liwei Jia
  10. Qiwei Li
  11. Guanghua Xiao
  12. Xiaowei Zhan

Reviewed by

Read the full article See related articles

Discuss this preprint

Start a discussion What are Sciety discussions?

Listed in

Log in to save this article

Abstract

Motivation

Spatial transcriptomics (ST) enables a high-resolution interrogation of molecular characteristics within specific spatial contexts and tissue morphology. Despite its potential, visualization of ST data is a challenging task due to the complexities in handling, sharing and visualizing large image datasets together with molecular information.

Results

We introduce ScopeViewer, a browser-based software designed to overcome these challenges. ScopeViewer offers the following functionalities: (1) It visualizes large image data and associated annotations at various zoom levels, allowing for intricate exploration of the data; (2) It enables dual interactive viewing of the original images along with their annotations, providing a comprehensive understanding of the context; (3) It displays spatial molecular features with optimized bandwidth, ensuring a smooth user experience; and (4) It bolsters data security by circumventing data transfers.

Availability

ScopeViewer is available at: https://datacommons.swmed.edu/scopeviewer

Contact

Xiaowei.Zhan@UTSouthwestern.edu , Guanghua.Xiao@UTSouthwestern.edu

Supplementary information

Supplementary data are available at Bioinformatics online.

Article activity feed

  1. GigaScience

    AbstractMotivation Spatial transcriptomics (ST) enables a high-resolution interrogation of molecular characteristics within specific spatial contexts and tissue morphology. Despite its potential, visualization of ST data is a challenging task due to the complexities in handling, sharing and visualizing large image datasets together with molecular information.Results We introduce ScopeViewer, a browser-based software designed to overcome these challenges. ScopeViewer offers the following functionalities: (1) It visualizes large image data and associated annotations at various zoom levels, allowing for intricate exploration of the data; (2) It enables dual interactive viewing of the original images along with their annotations, providing a comprehensive understanding of the context; (3) It displays spatial molecular features with optimized bandwidth, ensuring a smooth user experience; and (4) It bolsters data security by circumventing data transfers.

    This work has been peer reviewed in GigaScience (see https://doi.org/10.1093/gigascience/giag074), which carries out single-anonymized peer review. These reviews are published under a CC-BY 4.0 license and were as follows:

    (Resubmitted manuscript) Reviewer 3：

    This manuscript introduces ScopeViewer, a browser-based platform for visualizing large biological images, including spatial transcriptomics and pathology data. The topic is timely, and the software appears to be a more accessible and technically advanced implementation compared with existing visualization tools such as Loupe Browser or CellXGene Viewer. The work addresses an important technical challenge and presents a promising browser-based solution.

    However, to make the tool truly useful to the research community and to clearly distinguish it from existing tools, the manuscript still lacks essential information in two main aspects: biological relevance and usability for end users.

    Major Comments

    1. Biological Interpretation The manuscript should more clearly demonstrate how ScopeViewer contributes to biological discovery and interpretation. The current example focuses on cell segmentation and classification results, which are valuable but represent only a modest improvement over existing tools. The authors should elaborate on what additional types of biological information can be visualized or analyzed within the tool. For example:
    • What kinds of annotations are supported (e.g., cell boundaries, morphology, clusters, tissue regions, spot-level measurements)?
    • How are these annotations generated, formatted, and loaded into ScopeViewer?
    • How do these annotations facilitate the integration of molecular and cellular features for biological interpretation? I recommend providing an additional example with more specific biological questions, either in the supplementary materials or on the project website to demonstrate more diverse and distinctive applications of ScopeViewer. This would better highlight its scientific and practical advantages.
    1. Ease of Use and Reproducibility The usability of ScopeViewer could be enhanced through more detailed, step-by-step guidance. I strongly recommend including:
    • A reproducible and downloadable example dataset (with corresponding annotations) so users can easily replicate the workflow and apply it to their own data.
    • A clear data preparation guide describing how users can prepare or convert their datasets (for example, a typical 10x Visium dataset from a public repository), including what kinds of annotations can be generated and how they can be converted into the required JSON format. Providing this information in the documentation or supplementary materials would substantially improve accessibility, reproducibility, and community adoption.

    Minor Comment It would be helpful to clarify whether ScopeViewer supports newer high-resolution spatial omics platforms such as Visium HD, or Xenium, and to specify the extent to which users can visualize or explore data from these technologies.

    Overall Evaluation: ScopeViewer is a technically solid and promising tool with clear potential for broad application. Enhancing the manuscript with richer biological examples and comprehensive user guidance would significantly improve its impact and usability for the research community.

  2. GigaScience

    AbstractMotivation Spatial transcriptomics (ST) enables a high-resolution interrogation of molecular characteristics within specific spatial contexts and tissue morphology. Despite its potential, visualization of ST data is a challenging task due to the complexities in handling, sharing and visualizing large image datasets together with molecular information.Results We introduce ScopeViewer, a browser-based software designed to overcome these challenges. ScopeViewer offers the following functionalities: (1) It visualizes large image data and associated annotations at various zoom levels, allowing for intricate exploration of the data; (2) It enables dual interactive viewing of the original images along with their annotations, providing a comprehensive understanding of the context; (3) It displays spatial molecular features with optimized bandwidth, ensuring a smooth user experience; and (4) It bolsters data security by circumventing data transfers.

    This work has been peer reviewed in GigaScience (see https://doi.org/10.1093/gigascience/giag074), which carries out single-anonymized peer review. These reviews are published under a CC-BY 4.0 license and were as follows:

    (Resubmitted manuscript) Reviewer 2：

    In this manuscript, the authors present a web browser for visualizing and annotating histological or spatial omics data. While the work is highly needed, the usage and functions of the current version is quite limited. Because multiple providers of spatial transcriptomics technologies also present browser for histological and spatial transcriptomics data, the current web browser should be compared with those available browsers for clarifying the functionalities. In addition, an offline version should be provided to facilitate local applications. A detailed protocol/video should also be provided to help readers/users use the browser.

  3. GigaScience

    AbstractMotivation Spatial transcriptomics (ST) enables a high-resolution interrogation of molecular characteristics within specific spatial contexts and tissue morphology. Despite its potential, visualization of ST data is a challenging task due to the complexities in handling, sharing and visualizing large image datasets together with molecular information.Results We introduce ScopeViewer, a browser-based software designed to overcome these challenges. ScopeViewer offers the following functionalities: (1) It visualizes large image data and associated annotations at various zoom levels, allowing for intricate exploration of the data; (2) It enables dual interactive viewing of the original images along with their annotations, providing a comprehensive understanding of the context; (3) It displays spatial molecular features with optimized bandwidth, ensuring a smooth user experience; and (4) It bolsters data security by circumventing data transfers.

    This work has been peer reviewed in GigaScience (see https://doi.org/10.1093/gigascience/giag074), which carries out single-anonymized peer review. These reviews are published under a CC-BY 4.0 license and were as follows:

    (Resubmitted manuscript) Reviewer 1：

    ScopeViewer runs within a browser, eliminating the hassle of configuring specific Python or other environments required by similar tools. This browser-based cross-platform tool offers superior compatibility. It enables the dual-view comparison of annotated images and raw HE images, offering greater convenience than other tools. Additionally, ST files are typically large, and this tool significantly mitigates slow transfer issues for large files. Overall this is a useful tool that will be very helpful to the ST research community. However, this manuscript still has some issues that need to be addressed.

    1. Benchmarking against other tools shows that, although ScopeViewer offers the unique advantage of dual synchronised views, its overall performance is no better than that of other available tools. This distinctive strength should also be explicitly highlighted in the main text.
    2. While the web interface slightly lowers the barrier to entry for users, it may not offer revolutionary convenience. Many web-based tools allow users to drag and drop files or make simple selections, while ScopeViewer requires users to create a dedicated JSON file, which may be challenging for those with limited bioinformatics experience.
    3. I noticed some open-source code was missing. GitHub only hosts a subset of the project sources, not the complete repository.
    4. The manuscript claims that "ScopeViewer is the only tool that supports the browser native SQLite data format". Is there any literature or evidence substantiating this "uniqueness"?
    5. The application example provided in the article uses breast cancer data from the 10x Genomics platform only. Using only this single example, without data from other cancer types or measurement platforms, presents certain limitations.
  4. GigaScience

    AbstractMotivation Spatial transcriptomics (ST) enables a high-resolution interrogation of molecular characteristics within specific spatial contexts and tissue morphology. Despite its potential, visualization of ST data is a challenging task due to the complexities in handling, sharing and visualizing large image datasets together with molecular information.Results We introduce ScopeViewer, a browser-based software designed to overcome these challenges. ScopeViewer offers the following functionalities: (1) It visualizes large image data and associated annotations at various zoom levels, allowing for intricate exploration of the data; (2) It enables dual interactive viewing of the original images along with their annotations, providing a comprehensive understanding of the context; (3) It displays spatial molecular features with optimized bandwidth, ensuring a smooth user experience; and (4) It bolsters data security by circumventing data transfers.

    This work has been peer reviewed in GigaScience (see https://doi.org/10.1093/gigascience/giag074), which carries out single-anonymized peer review. These reviews are published under a CC-BY 4.0 license and were as follows:

    (Previous submission) Reviewer 2：

    General comment:

    The author delivers an interactive web server named ScopeViewer and a standalone Docker image for spatial transcriptomics data and histology data visualization. The server is featured by 1. enabling to adjust the view on an image; 2. Enable a co-visualization of histology images, annotations, and gene expression; 3. Display with optimized bandwidth; 4. Security of data transferring. However, although the co-visualization of histology and gene expression is one key step in annotating histological features and exploring gene expression, the study may be limited by 1. Lacking the unique value of ScopeViewer to the spatial biology; 2. Inadequately data support the full functions of ScopeViewer; 3, unclear writings. The contents below are my comments on the study and ScopeViewer.

    1. The uniqueness of ScopeViewer is not clear. Co-visualization of gene expression and histology images and other functionalities provided by ScopeViewer can be done by Loupe Browser. What were the downstream analyses enabled by ScopeViewer? What biological insight can be brought by the server? Can ScopeViewer be used for other high-resolution spatial transcriptomics data?
    2. There are several visualization methods for spatial transcriptomics data. For example, RESEPT, SpatialPCA, Vesalius, and SODB use pseudo-RGB images (e.g., transformed from gene expression) to visualize tissue heterogeneity. Would the author also adopt those methods for visualizing spatial transcriptomics data?
    3. Why did the author use HD-yolo for cell segmentation? Is there any benchmarking work to show HD-yolo is the best tool? Did the author consider additional options for providing the cell segmentation tool? In addition, is there any recommended histology image resolution, or what is the minimum resolution requirement for the current cell segmentation?
    4. The paper should fully discuss and demonstrate the functionalities of the server. Figures 1A, B, and C demonstrate the partial view of the server. The author should display a full view of the webpage regarding displaying main functions and corresponding explanations.
    5. ST is defined in the introduction, but its full name is still shown in the other part of the paper. Please be consistent in using abbreviations.
    6. Be specific about using molecular features. For this study, the author mainly focuses on spatial transcriptomics. Therefore, using gene expression is better than molecular features regarding the scope of transcriptomics unless the author can include other omics features in this study.
    7. Regarding "consider a standard pathology image of 20,000 by 20,000 pixels," please show the information on pixel size. For example, what is the physical length of each pixel?
    8. The display of example data may have some issues using the Safari browser. For example, in "Image Viewer" button of the "Examples" menu, the zoomed-in image cannot be zoomed out.
    9. In the introduction, "current software packages" is unclear. Please specify the package name.
  5. GigaScience

    AbstractMotivation Spatial transcriptomics (ST) enables a high-resolution interrogation of molecular characteristics within specific spatial contexts and tissue morphology. Despite its potential, visualization of ST data is a challenging task due to the complexities in handling, sharing and visualizing large image datasets together with molecular information.Results We introduce ScopeViewer, a browser-based software designed to overcome these challenges. ScopeViewer offers the following functionalities: (1) It visualizes large image data and associated annotations at various zoom levels, allowing for intricate exploration of the data; (2) It enables dual interactive viewing of the original images along with their annotations, providing a comprehensive understanding of the context; (3) It displays spatial molecular features with optimized bandwidth, ensuring a smooth user experience; and (4) It bolsters data security by circumventing data transfers.

    This work has been peer reviewed in GigaScience (see https://doi.org/10.1093/gigascience/giag074), which carries out single-anonymized peer review. These reviews are published under a CC-BY 4.0 license and were as follows:

    (Previous submission) Reviewer 1：

    The authors present a web-based tool for visualizing whole slide images, spatial transcriptomics and annotations. The tool is based on the OpenSeadragon viewer and is implemented in JavaScript using ReactJS. The data (images, markers, annotations) needs to be pre-processed to fit the format used in the software (dzi, sqlite, json respectively).

    In Section 2.1, the authors claim to support multiple annotation formats, but the content primarily discusses image formats. It's crucial for the authors to clarify this discrepancy, as their software currently only supports one annotation format—specifically: JSON files generated by the OpenSeadragon Annotorious plugin (not cited in the paper). The software also supports a single image format (DZI) and a single marker format (SQLite generated from CSV files), and none of these formats are annotation formats. Clearing up this terminology will enhance the precision of their software description.

    The authors introduce an interesting approach by employing an SQLite database to store spatial molecular data. However, given the current trend in the community towards utilizing the zarr format for efficient cloud access and computation (see SpatialData / NGFF), it is imperative that the authors provide a thorough comparison of the pros and cons of both formats. This will contribute to the relevance and adaptability of their software within the evolving landscape of spatial data storage.

    Comparative Analysis with Existing Tools To strengthen the manuscript, the authors should conduct a comprehensive comparison of their work with existing tools in the spatial transcriptomics community. This analysis should encompass both web-based tools such as Vitessce (https://dx.doi.org/10.31219/osf.io/y8thv), TissUUmaps (https://doi.org/10.1016/j.heliyon.2023.e15306), Cellxgene (https://doi.org/10.1101/2021.04.05.438318), Cytomine (https://doi.org/10.1002/prca.201800057) as well as non-web-based tools like Napari (https://doi.org/10.1017/S1431927622006328), Giotto (https://doi.org/10.1186/s13059-021-02286-2), ST viewer (https://doi.org/10.1093/bioinformatics/bty714), etc. Key aspects for comparison should include feature differences, speed benchmarks, and memory benchmarks. By conducting a thorough evaluation against established tools, the authors can highlight the unique features and advantages of their software. This will provide readers with a clearer understanding of how the proposed tool stands out in the current landscape.

    Additionally, it would be helpful if the authors could provide practical guidance on using their tool. I had difficulty finding information like the color scale for spatial transcriptomics data and how to retrieve spot values. The authors should include clear instructions or documentation explaining how to access these features within the tool.

    Given the feedback provided, it appears that the manuscript lacks the necessary innovation and fails to sufficiently distinguish itself in the context of existing tools in the spatial transcriptomics field. The suggested revisions, including emphasizing format distinctions, conducting a comparative analysis, and exploring the advantages and disadvantages of the chosen data storage format, are crucial for addressing these concerns. However, despite these proposed changes, the current state of the manuscript does not offer significant advancements or unique contributions that would set it apart from established tools. As a result, I recommend rejecting the paper due to its limited impact and relevance within the spatial transcriptomics community.

  6. Version published to 10.1101/2023.07.24.549256 on bioRxiv