I'd like to announce FindSSU, a new tool in BBTools 39.86 for identifying and classifying ribosomal SSU (16S/18S) sequences. It compares query sequences against a curated database of 276,772 organisms (221,000 prokaryotic 16S and 55,000 eukaryotic 18S) using a DynamicDemiLog index (upcoming paper) with alignment-based exact ANI calculation.

What it does:

• SSU classification: Input one or more 16S/18S sequences, get the closest database matches ranked by alignment ANI.
• Gene-calling mode: Input a genome or contig. FindSSU will call SSU genes internally and classify each one found.
• Sequence lookup: Retrieve the SSU sequence for any organism by name or NCBI TaxID. Accepts full names (Escherichia_coli), abbreviated genus.species (E.coli), or numeric TaxIDs (562).

Three ways to use it:

1. Web interface — paste sequences, upload a file, or look up organisms by name: https://bbmap.org/services/findssu

2. Command line via BBTools:

• Classify SSU sequences: findssu.sh mySSU.fa records=3
• Gene-call from a genome: findssu.sh genome.fa call
• Look up E. coli's 16S (currently requires local mode): findssu.sh name=E.coli local (Note - full name gives better disambiguation - 'name=Escherichia_coli')
• Run locally (no network, requires resource files): findssu.sh mySSU.fa local

1. API — POST to the public endpoint for programmatic access:

• Classify a sequence (JSON response): curl -X POST https://bbmapservers.jgi.doe.gov/sendclade/findssu/ -H 'Content-Type: text/plain' -d $'//JSON\n//records=3\n//lineage=t\n>my_query\nGATGAACGCTGGCGGCAGGCCTAACACATGCAAGTCG...'
• Look up by name: curl -X POST https://bbmapservers.jgi.doe.gov/sendclade/findssu/ -H 'Content-Type: text/plain' -d $'//JSON\n//lineage=t\n//name=Meiothermus_ruber'
• Look up by TaxID: curl -X POST https://bbmapservers.jgi.doe.gov/sendclade/findssu/ -H 'Content-Type: text/plain' -d $'//JSON\n//tid=562'

Output includes alignment ANI, weighted kmer identity (WKID), SSU type (16S/18S), sequence lengths, NCBI TaxID, organism name, contig name/strand/start position, and optional full taxonomic lineage. JSON and tab-delimited formats supported.

Database: The reference database contains 276,772 SSU sequences derived from both Silva and gene-calling on NCBI RefSeq, with broader taxonomic coverage than either alone. The optional resource files can be downloaded separately from SourceForge - https://sourceforge.net/projects/bbmap/files/Resources/ - ssuSketchDDL.tsv.gz, all_prok_16S_best_taxsorted.fa.gz, and all_euk_18S_best_taxsorted.fa.gz. Roughly 90MB total; just drop them into BBTools/resources/

Performance: Classification against the full database typically completes in under a second plus milliseconds per additional query via the server, or thousands of queries per second locally. The tool uses an inverted index to reduce the search space to a few dozen candidates per query, followed by SSU alignment for exact ANI. Running the full database locally uses <1GB RAM.

Future: I plan to add 5S and 23S, and possibly 5.8S, 28S, and fungal ITS sequences!