Example: download and search 66,000 GTDB genomes with a query genome
sourmash-bio/sourmash-examples#13
You'll need to build the genome signature file in Example: create a signature by downloading and sketching a genome sequence first.
Then, download the GTDB genomic representatives database:
curl -JLO [https://osf.io/3a6gn/download](https://osf.io/3a6gn/download)
This will create a 1.7 GB file, gtdb-rs207.genomic-reps.dna.k31.zip, which contains 66,000 genome sketches from the Genome Taxonomy Database, release 207.
Now search the genome against the GTDB database:
sourmash search GCF_000005845.2_ASM584v2_genomic.fna.gz.sig gtdb-rs207.genomic-reps.dna.k31.zip
This will take about 5 minutes.
The output will look like this:
8 matches; showing first 3:
similarity match
---------- -----
29.9% GCF_003697165.2 Escherichia coli DSM 30083 = JCM 1649 = A...
14.6% GCF_002965065.1 Escherichia sp. MOD1-EC7003 strain=MOD1-E...
14.2% GCF_000026225.1 Escherichia fergusonii ATCC 35469 strain=...
showing that this genome is, indeed, an E. coli genome :).
The similarity in the left column is Jaccard similarity, calculated using the k-mers in the query genome sketch against the k-mers in each of the database genome sketches.
You can increase the number of output results with -n:
8 matches:
similarity match
---------- -----
29.9% GCF_003697165.2 Escherichia coli DSM 30083 = JCM 1649 = A...
14.6% GCF_002965065.1 Escherichia sp. MOD1-EC7003 strain=MOD1-E...
14.2% GCF_000026225.1 Escherichia fergusonii ATCC 35469 strain=...
14.1% GCF_902498915.1 Escherichia ruysiae, OPT1704
14.1% GCF_004211955.1 Escherichia sp. E1V33 strain=E1V33, ASM42...
13.5% GCF_005843885.1 Escherichia sp. E4742 strain=E4742, ASM58...
10.3% GCF_001660175.1 Escherichia sp. B1147 strain=B1147, ASM16...
10.1% GCF_011881725.1 Escherichia coli strain=SCPM-O-B-8794, AS...
and you can record the results in a CSV file with -o <output.csv>.
Categories
This example belongs to the following categories: