Contribute to STAr

Spatial Transcriptomics Arena (STAr) is a community resource for spatially resolved transcriptomics datasets, spatially variable gene (SVG) detection methods, and reproducible analysis results. We welcome contributions that are publicly shareable, adequately documented, and suitable for integration into the STAr data structure.

Submit Your Contribution

The website does not currently provide direct user upload. To propose a contribution, use the Contact page with one of these subject lines:

STAr contribution: dataset

STAr contribution: method

STAr contribution: analysis results

Provide stable public links to files, code, and publication. Do not send restricted or personally identifiable data.

Contribution Workflow

Submit a Proposal

Send your contribution information through the Contact page with required documentation.

Initial Screening

Maintainers verify relevance, accessibility, documentation, and STAr compatibility.

Quality Control

Verify file integrity, metadata completeness, identifier agreement, and reproducibility.

Revision if Needed

Clarify metadata, correct file-format problems, or supply missing information.

Integration & Release

Accepted resources are standardized and added during periodic website updates.

Contribution Types

You may contribute either a public repository accession or a processed data package.

Option 1: Public Repository Accession

Provide the following information:

  • Repository name and accession or permanent URL (GEO, Zenodo, Dryad, etc.)
  • Dataset title and citation or preprint
  • Organism, tissue, disease/experimental condition, and spatial technology
  • Sample, section, field-of-view, or replicate structure
  • Locations of expression matrix, spatial coordinates, and histology image
  • Brief description of preprocessing performed
  • Reuse license or data-use restrictions
Example Submission:
Contribution type: Dataset
Dataset: Mouse olfactory bulb generated with 10x Visium
Repository: GEO
Accession: GSM4656184
Organism: Mouse
Tissue: Olfactory bulb
Condition: Not applicable
Samples: One tissue section
Files: Count matrix, spot coordinates, and histology image
Preprocessing: Raw counts; no normalization applied
Citation: [DOI or publication URL]
License or restrictions: [license or repository terms]
Option 2: Processed Data Package

Provide a compressed archive containing the following files:

Required Files
  • Expression: Location-by-gene matrix or sparse format
  • Spatial coordinates: x, y coordinates with matching identifiers
  • Gene annotation: Gene ID and name/symbol
  • Image: Optional histology image with alignment info
  • README: File descriptions and preprocessing details
  • Metadata JSON: Dataset description record
Format Notes
  • CSV/TSV for moderate-sized datasets
  • Matrix Market (.mtx) for large sparse matrices
  • Identifiers must match across files
  • Document all preprocessing and normalization
  • Use null only when unavailable
  • Explain unavailable items in README
Minimum Metadata Fields (JSON):
{
  "dataset_name": "mouse_olfactory_bulb_example",
  "display_name": "Mouse Olfactory Bulb Example",
  "dataset_type": "Real Data",
  "technology": "10x Visium",
  "organism": "Mouse",
  "tissue": "Brain - olfactory bulb",
  "disease": null,
  "raw_data": ["expression.mtx", "features.tsv", "barcodes.tsv", "coordinates.csv"],
  "processed_data": ["processed_dataset.zip"],
  "normalized_expression": [],
  "number_of_genes": 20000,
  "number_of_location": 3000,
  "zero_percentage": 92.5,
  "source_name": "Author_Year",
  "source_link": "https://doi.org/...",
  "image_file": ["histology_image.png"]
}

The method must be publicly accessible with a runnable example that accepts a spatial expression dataset and produces gene-level SVG statistics.

Required Information
  • Method name, citation/preprint, and description
  • Public source-code or package URL
  • Open-source license
  • Software version or commit identifier
  • Programming language and dependencies
  • Expected input files and preprocessing needs
  • Minimal command/script/vignette for example
  • All parameter settings and thresholds
  • Expected output columns and example file
  • Approximate runtime and memory requirements
Example Submission:
Contribution type: Method
Method: ExampleSVG
Code: https://github.com/example/example-svg
Version: v1.2.0
License: MIT
Language: R 4.5
Input: xpression matrix and x-y coordinate table
Example: Rscript run_example.R \
  --counts counts.csv \
  --coords coordinates.csv
Parameters: Default settings;
  adj-p threshold = 0.05;
  seed = 2026
Output: example_results.csv
  columns: gene_id, statistic,
  p_value, adjusted_p_value, rank
Citation: https://doi.org/...
Important: The example output should contain one row per gene. If the method doesn't produce a p-value, document the alternative evidence measure and how values should be interpreted.

Analysis results may be submitted when both the dataset and method are publicly documented. Provide the exact dataset version, method version, preprocessing, parameters, and significance rule used.

Required Result Columns
dataset_id Dataset identifier statistic Test statistic value
sample_id Sample identifier p_value Raw p-value
gene_id Unique gene ID adjusted_p_value Adjusted p-value
gene_name Gene symbol bayes_factor Bayes factor
method Method name rank Gene rank
method_version Method version is_svg Boolean (TRUE/FALSE)

At minimum, each result row must include the dataset and sample identifiers, gene identifier, method and version, and a documented statistic or rank. Use NA only when a quantity is not produced by the method; define every use of NA in the accompanying README. For is_svg, state the exact calling rule, such as an adjusted p-value below 0.05 or a specified Bayes-factor threshold.

Example Results Format:
dataset_id,sample_id,gene_id,gene_name,method,method_version,statistic,p_value,adjusted_p_value,bayes_factor,rank,is_svg
mouse_ob_example,section_1,ENSMUSG00000000001,GeneA,ExampleSVG,1.2.0,12.41,0.0002,0.004,NA,1,TRUE
mouse_ob_example,section_1,ENSMUSG00000000002,GeneB,ExampleSVG,1.2.0,4.18,0.1200,0.310,NA,152,FALSE
Additional Documentation Required
  • README describing dataset, method, preprocessing, parameters, and environment
  • Script or workflow used to generate the results
  • Number of genes tested and number called as SVGs (per sample)
  • Links to corresponding dataset and method (if not already in STAr)

Quality-Control Checklist

Before submitting, confirm that all items below are complete:

Questions? Contact us through the Contact page before preparing a full submission. We welcome questions about unusual formats and new technologies.