1.0.2

v1.0.2 (2026-02-03)

🐛 Bug Fixes

• Fix inconsistent phenotype_shared_annotations types (list[dict])
  Issue: #150
  Thanks @kinari-labwork for reporting the issue!!

• Update template paths in webapp_builder.py to reflect current directory structure.
  Issue: #154

• Use Jaccard index in _calculate_pair_worker for pairwise similarity computation.
  Issue: #156

🔧 Maintenance

• Rename pairwise_similarity_annotations key from "phenotype" to "mp_term_name".
  Issue: #151

• Automate E2E tests for the TSUMUGI subcommand using GitHub Actions
  Issue: #152

• Restrict tsumugi genes input to text files only for better maintainability.
  Issue: #153

• Require explicit --genewise or --pairwise option for tsumugi genes and mp commands.
  Issue: #155

1.0.1

v1.0.1 (2026-01-27)

🖼️ Web Interface

• https://larc-tsukuba.github.io/tsumugi/

🌟 New Features

• Add numeric inputs for viewer sliders and sync with controls
  Issue: #148

🐛 Bug Fixes

• Fix a bug of "No Gene Network Found” popup is not displayed when a gene has no network.
  Issue: #146
  Thanks to @aki2274 for reporting the issue!

• Fix tooltip hover from disabling network pan.
  Issue: #147

1.0.0

v1.0.0 (2026-01-20)

🌟 New Features

• Add draggable module frames and tooltip helpers for Cytoscape subnetwork grouping.
  Issue: #125
  [Commit Detail]

• Add SVG export with configurable scale in the viewer.
  Issue: #127
  [Commit Detail]

• Add toggle controls for full-screen mode and responsive Cytoscape resizing.
  Issue: #128
  [Commit Detail]

• Enable command-line filters for pairwise and genewise modes.
  Issue: #129
  [Commit Detail]

• Add --pairwise option in tsumugi genes.
  Issue: #133
  [Commit Detail]

🖼️ User Interface

• Make Cytoscape tooltips resizable with scrollable sections.
  Issue: #130
  [Commit Detail]

• Scope neighbor dimming to the clicked module and refine node highlight emphasis.
  Issue: #134
  [Commit Detail]

• UI and interaction improvements for network visualization.
  Issue: #136
  [Commit Detail]
  Thanks to @aki2274 for reporting this issue!

• Move phenotype scale SVGs below the graph and center them with updated sizing.
  Issue: #137
  [Commit Detail]

• Hide phenotype severity scale and legend for binary phenotypes.
  Issue: #138
  [Commit Detail]

• Add All handling and a two-column layout for biological annotation filters.
  Issue: #139
  [Commit Detail]
  Thanks to @kinari-labwork for reporting to this issue!

🐛 Bug Fixes

• Fix node count inconsistency after resetting biological annotation filters.
  Issue: #141
  [Commit Detail]

• Preserve edges when resetting genotype filters to All.
  [Commit Detail]

• Fix version file output directory to include the app subdirectory.
  [Commit Detail]

• Fix an issue where selected phenotypes remained in the search results even after being excluded by the Biological annotation filter.
  Issue: #142
  [Commit Detail]

• Fix an issue where an isolated node unexpectedly remained in a phenotype JSON.gz file
  Issue: #143
  [Commit Detail]

🚀 Performance

• Codebase refactoring for improved modularity and memory efficiency.
  Issue: #140
  [Commit Detail]

• Refactor similarity calculators to single-threaded, flattened lookups.
  [Commit Detail]

• Stream JSONL outputs as iterators to reduce memory usage.
  [Commit Detail]

• Use gzip compression level 9 for smaller outputs.
  [Commit Detail]

📝 Documentation

• Document CLI filters for pairwise and genewise modes.
  [Commit Detail]

• Update MP term filtering documentation in multiple languages.
  [Commit Detail]

• Update 'Download Data' Section on TSUMUGI Top Page.
  Issue: #132

⚙ Maintenance

• Generate phenotype/gene HTML and JS on the client side for web builds.
  Issue: #124
  [Commit Detail]

0.5.0

v0.5.0 (2025-11-25)

🌟 New Features

• CLI tooling suite added for offline/pipe-friendly workflows:
  • tsumugi run (main pipeline)
  • tsumugi mp --include/--exclude
  • tsumugi n-phenos --pairwise --min --max
  • tsumugi n-phenos --genewise --min --max
  • tsumugi genes --keep --drop
  • tsumugi life-stage --keep --drop
  • tsumugi sex --keep --drop
  • tsumugi zygosity --keep --drop
  • tsumugi build-webapp
  • tsumugi build-graphml

📝 Documentation

• README updated with CLI option descriptions and multilingual links (Japanese, Korean, Simplified/Traditional Chinese, Hindi, Indonesian, Vietnamese, Spanish, French, German, Portuguese).

0.4.0

💥 Breaking

• Adopted the Phenodigm/Resnik similarity pipeline so phenotype matching now respects the hierarchical structure of MPO instead of relying on plain Jaccard overlap.
  Issue: #71
  [Commit Detail]

• Surface common ancestor MP terms on KO mouse edges, filtering out low-information matches to highlight meaningful shared phenotypes.
  Issue: #118
  [Commit Detail]

• Pre-compute degree and betweenness centrality for every node and expose slider-based filters to spotlight highly connected genes.
  Issue: #108
  [Commit Detail]

🌟 New Features

• Expanded the phenotype similarity and severity scales from 0–10 to 1–100 so users get Phenodigm-level precision when scoring networks.
  Issue: #119
  [Commit Detail]

• Added a JPEG export mode alongside the existing PNG and SVG download options for Cytoscape.js views.
  Issue: #117
  [Commit Detail]

• Normalized degree and betweenness scores by phenotype count and introduced an optional log-scale transform to curb hub bias.
  Issue: #113
  [Commit Detail]

• Delivered Cytoscape Desktop–compatible GraphML exports (nodes and edges) for seamless round-tripping between the web app and local analyses.
  Issue: #92
  [Commit Detail]

0.3.2

v0.3.2 (2025-06-23)

💥 Breaking Changes

• Updated to support IMPC Release 23.0
  Upgraded the underlying dataset to the latest IMPC Release 23.0 for improved data coverage and accuracy.
  Issue: #88
  [Commit Detail]

🌟 New Features

• Human disease gene highlighting
  Added functionality to highlight genes associated with human diseases, enabling researchers to identify disease-relevant gene modules within phenotypic networks. The feature uses Phenodigm algorithm to establish mouse-human disease relationships.
  Issue: #87
  [Commit Detail]
  Thanks to @kinari-labwork for contributing to this issue!

🖼️ User Interface

• Enhanced footer layout
  Redesigned the footer to better emphasize documentation and inquiry options, improving user access to help resources.
  Issue: #98
  [Commit Detail]

• Unified mobile control panel
  Consolidated the previously split left/right control panels into a single, unified panel view on mobile devices for improved usability and screen space utilization.
  Issue: #89
  [Commit Detail]

• Dynamic header font sizing
  Implemented responsive header font sizing that automatically adjusts on mobile devices to ensure text displays properly within a single line, preventing layout issues.
  Issue: #93
  [Commit Detail]

🐛 Bug Fixes

• Fixed mobile toggle button visibility
  Resolved an issue where the toggle close button was not appearing on mobile devices, restoring proper navigation functionality.
  Issue: #94
  [Commit Detail]

• Optimized network rendering performance
  Implemented a performance optimization by limiting the maximum number of displayed nodes to 150 (reduced from 200) to prevent rendering delays and ensure smooth user experience with large networks.
  Issue: #103
  [Commit Detail]

• Improved gene symbol page visualization
  Modified gene symbol pages to display only nodes that are directly connected to the target gene, reducing visual clutter and focusing on relevant gene relationships.
  Issue: #104
  [Commit Detail]

🔧 Maintenance

• Enhanced raw data distribution
  Added version tags to TSUMUGI_raw_data filenames and expanded format support to include both CSV and Parquet formats, improving data accessibility and compatibility with different analysis tools.
  Issues: #95, #96
  [Commit Detail]

• Streamlined deployment workflow
  Consolidated the separate testing and production deployment notebooks into a single, unified deployment file, simplifying the release process and reducing maintenance overhead.
  Issue: #101
  [Commit Detail]

0.3.1

🖼️ User Interface

• The user-specified phenotype is now always displayed, regardless of filter settings
  Issue #80
  [Commit Detail]

• Added TSUMUGI version information and a contact form to the footer
  Issue #79
  [Commit Detail]

🐛 Bug Fixes

• Ensured that all nodes and edges related to the target phenotype are properly displayed
  Issue #54
  [Commit Detail]

• To prevent nodes from disappearing when adjusting the slider, the row with the largest absolute effect size (between Homo and Hetero) is now selected
  Issue #72
  [Commit Detail]

• edgeMin is now defined as the minimum edge size among edges connected to the target gene, similar to the edgeMax logic
  Issue #74
  [Commit Detail]

• Nodes that do not contain the user-specified phenotype are now excluded from display
  Issue #54
  [Commit Detail]

• Only edges between nodes that share two or more phenotypes are now visualized
  Issue #81
  [Commit Detail]

• Fixed an issue where duplicate phenotypes were shown in tooltips
  Issue #82
  [Commit Detail]

🔧 Maintenance

• Improved visibility in gene search by automatically centering and zooming in on the corresponding node
  Issue #78
  [Commit Detail]
  Thanks for your comments, @aki2274!

0.3.0

💥 Breaking

• Annotate the four life stages: Embryo, Early, Interval, and Late. Issue #20.

  • [Commit Detail]
  • [Commit Detail]

• Set thresholds for visualization considering similarity. Issue #77. [Commit Detail].

  • Thanks to @aki2274 for reporting this issue!

• Rename the header name cluster to module in the exported CSV. [Commit Detail].

📝 Documentation

• Add documentation on how to load the last column (List of shared phenotypes) in TSUMUGI_raw_data.csv.gz as list[str]. [Commit Detail].

• Add documentation of life stage. [Commit Detail].

🖼️ User Interface

• Organize the control panel. [Commit Detail]

0.2.2

🖼️ User Interface

• Add support for smartphone operation.
  Commit Detail

• Redesign the control panel for improved usability.
  Issue: #65
  Commit Detail

📝 Documentation

• Publish a complete list of currently available phenotypes and genes.
  Issue: #34
  Commit Detail

🔧 Maintenance

• Use inherit for tab button text color to ensure consistent appearance and prevent default blue styling on macOS.
  Issue: #67
  Commit Detail

• Unify HTML templates for better maintainability.
  Issue: #68
  Commit Detail

• Unify JavaScript templates to reduce redundancy.
  Issue: #69
  Commit Detail

🐛 Bug Fixes

• Adjust the severity slider behavior to always keep at least one gene pair visible, even at minimum or maximum values.
  Issue: #72
  Commit Detail

• Change the link button font color from blue to black (#333) on macOS for consistent styling.
  Issue: #67
  Commit Detail

0.2.1

📝 Documentation

• Add documentation (README.md and README_JP.md). Issue #60 [Commit Detail]

🌟 New Features

• Add a Google Form to allow anyone to submit inquiries. [Commit Detail]

• Add a search box for a gene symbol in the network Issue. #30 [Commit Detail]

🐛 Bug Fixes

• Modified to ensure that the Submit button is definitely disabled when the input is empty. Issue: #61 [Commit Detail]

• Phenotypes for which no phenotype-similar gene groups were detected have been removed from available_mp_terms.json to prevent them from being selected. Issue: #58 [Commit Detail]

• Fixed a bug where target genes disappeared when increasing the slider threshold in "Gene". Issue: #63 [Commit Detail]

🖼️ User Interface

• Improve the visibility of the top page tab. Issue: #59 [Commit Detail]