metapang search pangenome

Match a query against a single species pangenome de Bruijn graph to find which gene families it contains. This is the read-mapping step of metapang profile exposed on its own, for one pangenome.

Unlike metapang search bank, which compares MinHash sketches across a whole collection, this command uses metagraph to look up exact k-mers in one pangenome graph and report, for each query sequence, the gene-family labels it matches.

Synopsis

metapang search pangenome -q query.fasta.gz -b GTDB_refseq@2.0.0:s__Abiotrophia_defectiva [OPTIONS]
metapang search pangenome -q query.fasta.gz -b local:my_pangenome                          [OPTIONS]
metapang search pangenome -q query.fasta.gz -g pangenome.dbg -a pangenome.annodbg          [OPTIONS]

Inputs

The query is required, and the pangenome is given in one of three ways: from PanGBank, a local pangenome built with metapang cache add, or an explicit local de Bruijn graph plus its annotations.

-q, --query (required)

The query file (fasta/fastq, gzipped accepted), a genome or reads.

-b, --pangbank

Use a pangenome from PanGBank, as collection[@version]:name (for example GTDB_refseq@2.0.0:s__Abiotrophia_defectiva, downloaded and cached on first use), or a local pangenome as local:<name> built with metapang cache add (see Local pangenomes).

-g, --dbg

A local pangenome de Bruijn graph (the .dbg file).

-a, --annotations

The annotations for that graph (the .annodbg file).

Note

Give either --pangbank (a PanGBank or local: pangenome), or both --dbg and --annotations. The --dbg / --annotations pair is the per-pangenome output of metapang index bank.

Options

Option	Default	Description
-t, --threads	1	Threads used by metagraph for the query.
--metagraph-path	metagraph	Path to the metagraph binary.
-o, --output	{pangenome_name}_{query_file_name}.tsv	Output file. Use stdout or - to write to standard output.
--annotate		Also write a graph_tool annotated pangenome graph (.gt) to this path, with per-family read and k-mer counts, as metapang profile produces.

Output

A TSV produced by metagraph: for each query sequence, the pangenome gene families (labels) whose k-mers it matches. This is the same per-family mapping that profile uses internally to derive gene-family coverage.

Annotated graph (--annotate)

With --annotate graph.gt, the query result is projected back onto the pangenome graph: the prebuilt .gt for the species is loaded and each gene family is annotated with its read and k-mer counts, then saved to the given path. This is the same annotated graph metapang profile builds internally.

Note

--annotate requires --pangbank (a PanGBank or local: pangenome) It also needs the query results in a file, so it cannot be combined with dout`.

Examples

# search a genome against one PanGBank pangenome
metapang search pangenome -q genome.fasta.gz \
  -b GTDB_refseq@2.0.0:s__Abiotrophia_defectiva -t 8

# search against a local pangenome (built with 'metapang cache add')
metapang search pangenome -q genome.fasta.gz -b local:my_pangenome -o -

# search an explicit dbg / annotations pair and print to stdout
metapang search pangenome -q genome.fasta.gz \
  -g s__Abiotrophia_defectiva.dbg -a s__Abiotrophia_defectiva.annodbg -o -