Skip to content

Latest commit

 

History

History
47 lines (32 loc) · 1.63 KB

make-sequence-table.md

File metadata and controls

47 lines (32 loc) · 1.63 KB

make-sequence-table

Build a sample-by-sequence feature table. Mirrors R's makeSequenceTable().

dada2-rs make-sequence-table dada/*.json -o seqtab.json

Reads JSON produced by dada (one file per sample) or merge-pairs (one file containing multiple samples) and assembles a flat count matrix, samples × sequences.

Input

<INPUT>... — dada or merge-pairs JSON files.

--sample-names — comma-separated sample names, one per input file, in input order. Only applies to single-sample dada files; merge-pairs files carry sample names internally. If provided, the length must match the number of inputs.

Filtering

--min-len / --max-len — discard ASV sequences shorter / longer than this length, inclusive. The usual reason to set these is removing off-target amplicons, which show up as a length mode well away from the expected amplicon size.

Output

--order-by (default abundance) — column order: abundance (decreasing total), nsamples (decreasing number of samples present in), or none (first-seen order).

--hash (default md5) — hash algorithm for sequence identifiers; md5 or sha1. MD5 is the default for interoperability: it is what QIIME 2 and other tools in the ecosystem use, so the same sequence gets the same identifier across them. The choice is about ecosystem consistency, not cryptographic strength.

--output / -o — write JSON here instead of stdout.

--compact — minified JSON.

See also