Metadata from delimited files

In all seqtool commands, it is possible to integrate external metadata from delimited text files created manually or using another program.

Files are specified using the -m/--meta option and accessed using the functions meta(column), opt_meta(column) (with missing data) or has_meta(column) (to check if the metadata is present). Column is either a number or the header name of the given column.

See also variable reference and detailed description of command-line options

By default, files are assumed to be tab-delimited, and the first column should contain the ID. However, this can be changed with --meta-delim and --id-col.

Examples

Consider this list containing taxonomic information about sequences (genus.tsv):

id  genus
seq1  Actinomyces
seq2  Amycolatopsis
(...)

The genus name can be added to the FASTA header using this command:

st set --meta genus.tsv --desc '{meta(genus)}' input.fasta > with_genus.fasta
# short:
st set -m genus.tsv -d '{meta(genus)}' input.fasta > with_genus.fasta

>seq1 Actinomyces
SEQUENCE
>seq2 Amycolatopsis
SEQUENCE
(...)

If any of the sequence IDs is not found in the metadata, there will be an error. If missing data is expected, use opt_meta instead. Missing entries are undefined:

st set -m genus.tsv --desc '{opt_meta(genus)}' input.fasta > with_genus.fasta

>seq1 Actinomyces
SEQUENCE
>seq2 Amycolatopsis
SEQUENCE
>seq3 undefined
SEQUENCE
(...)

Filtering by ID

Sometimes it is necessary to select all sequence records present in a list of sequence IDs. This can easily be achieved using this command:

st filter -m id_list.txt 'has_meta()' seqs.fasta > in_list.fasta

Multiple metadata sources

Several sources can be simultaneously used in the same command with -m file1 -m file2 -m file3...:

st filter -m source1.txt -m source2.txt 'meta("column", 1) == "value" && has_meta(2)' seqs.fasta > in_list.fasta

Sources are referenced using meta(column, file_number) or has_meta(file_number); see also variable reference