Skip to content
Documentation
Tools
Sequence Similarity Search

Sequence Similarity Search

Sequence Similarity Search is available at Sequence Similarity Search. It runs minimap2 in the visitor’s browser through Aioli/Biowasm and does not use a CapBase backend server.

User Guide

Go to Sequence Similarity Search > Paste nucleotide FASTA or upload a FASTA file > Choose a Capsella genome > Choose a reference database > Click Run Search > Review the parsed PAF results.

Searches run locally in the browser. Query sequences are not uploaded to CapBase.

If the browser cannot fetch a GitHub Release .mmi index directly, use the Download selected index link on the tool page, then choose the downloaded .mmi file in the Local minimap2 index file field and run the search again. The local file name must match the selected database.

Databases

  • Genome: complete genome assembly, including intergenic regions.
  • Gene genomic regions: genomic gene sequences that may include introns.
  • Transcripts: annotated spliced transcript sequences.
  • Coding sequences: annotated CDS sequences.
  • Transposable elements: annotated TE copies.

Developer Notes

The tool page is defined at:

docs/content/sequence-search/index.md
layouts/shortcodes/sequence-similarity-search.html

The browser-side code is split into modules under:

static/js/sequence-search/

The static database configuration is:

static/sequence_search/databases.json
docs/static/sequence_search/databases.json

The minimap2 indexes are large .mmi files hosted as GitHub Release assets:

https://github.com/DrChenHeng/CapBase/releases/tag/sequence-search-v1

Do not commit .mmi files to the repository. The local folder static/gene_cards/indexes/ is for local testing or staging only and is ignored by Git. To update a database, build the replacement index, upload it to the sequence-search-v1 GitHub Release, record its byte size, and update both database configuration files.

Example upload command:

gh release upload sequence-search-v1 \
  Crubella_transcriptome.mmi \
  --repo DrChenHeng/CapBase \
  --clobber

If a new species is added, add its Release-hosted .mmi URLs, FASTA paths, .fai paths, .gzi paths, preset, and record-link type to the database configuration.

Index Building

Example minimap2 index command:

minimap2 -d Crubella_transcriptome.mmi \
  Crubella_RAM_transcriptome_v3.2.fa.gz

The .fai and .gzi files support indexed sequence retrieval, but they do not replace .mmi files for similarity searching.

The indexes in sequence-search-v1 were generated for minimap2 2.22, matching the Biowasm package configured as minimap2/2.22.

To test Release asset redirects and CORS, open the Sequence Similarity Search page through a local Hugo server, select Capsella rubella and Transcripts, and run a short nucleotide FASTA query. The browser network panel should show the selected .mmi downloading from the GitHub Release only after Run Search is pressed. If the browser reports Failed to fetch, use the local .mmi fallback or move the indexes to CORS-enabled static hosting and update the index URLs in the database configuration.

Current Limitations

  • Nucleotide search only; protein FASTA files are not searched by this tool.
  • This is minimap2, not NCBI BLAST.
  • Large .mmi indexes can require substantial browser memory.
  • Desktop browsers are recommended.
  • The WebAssembly runtime is loaded from the pinned Biowasm CDN path https://biowasm.com/cdn/v3/aioli.js; minimap2 package version is configured as minimap2/2.22.
Ortholog FinderGenome Browser