BBTools QuickClade

QuickClade assigns taxonomy to query sequences by comparing k-mer frequencies to those in a reference database, supplemented by DDL sketch matching using k25 indexes for improved accuracy. It is extremely fast and uses little memory, making it ideal for taxonomic classification of genomic sequences including metagenomic bins. The reference database covers approximately 150,000 genomes from RefSeq for comprehensive taxonomic assignment.

Running QuickClade Locally

To run QuickClade locally (without the web server), place two files in the /resources/ directory of your BBTools installation:

The spectra file (e.g., refseqA48_with_ribo.spectra.gz) — the main reference database of k-mer frequency profiles
The DDL sketch file (e.g., refseqSketchDDL_k25e5b4096.tsv.gz) — for sketch-based matching using k25 indexes

Both files are auto-discovered from /resources/ when running quickclade.sh locally. If no local database is found, use the server flag to send k-mer spectra to a remote server instead.

Input Method:
Paste Sequences Upload File

Input Sequences (FASTA format):

Size: 0 MB / 15 MB limit

Processing Mode:
Java (sends raw sequence) Rust (sends kmer spectra)

Classification Mode:
Single Query (combine all sequences) Per Sequence (classify each individually)

Max Results: