This directory contains a Makefile that will build a private database for the 64 genomes used in Awad et al., 2017.
This example extends example.private with gene finding on the input genomes using prodigal.
Run make to run the entire pipeline. (You'll need sourmash v4.4.0
installed, along with snakemake >= 6 and prodigal.)
The Makefile does the following:
The Makefile runs curl to download the genomes from
the OSF project, and then unpacks them
into the podar-ref/ directory.
The Makefile will next run the snakemake workflow in Snakefile to
build _protein.faa files for all the .fa files in the podar-ref/
directory.
Next, the Makefile uses the script ../fasta-to-fromfile.py to scan
the genomes and proteomes and then produces a summary file,
build.csv, that contains names and source files for building
sourmash signatures.
Here, fasta-to-fromfile uses the identifiers present in the sequences
to connect the genome and proteome so that they are sketched with the same
names.
Finally, the Makefile runs
sourmash sketch fromfile build.csv -p dna -p protein -o podar-ref.zip
to sketch all of the genomes in build.csv. The parameter string -p dna tells sourmash to construct DNA sketches, and the parameter
string -p protein constructs protein sketches. Here, sourmash sketch fromfile automatically selects the genome for building the DNA
sketches and the proteome for building the protein sketches.
The names for the output signatures are taken from build.csv.
You can run sourmash sig summarize podar-ref.zip to get a summary of
the contents of the zip file, or sourmash sig describe podar-ref.zip
to get a listing of all the signatures.
You can get a detailed listing of just one pair of the signatures by using
the --include-db-pattern option for sig describe:
sourmash sig describe podar-ref.zip --include NZ_FWDH