Is there a way to sketch 1,000 files in one line

[ZeweiSong]

Hi,

I was wondering if there is an option that I can sketch many files in one command, I would also like to get all pairwise distances.

Thanks!

Zewei

[ondovb]

Hi Zewei,

Not built into Mash, but you can use some shell trickery (see #26):
echo a b c | xargs -n 1 mash sketch.

For the distances, you would then need to paste them and give the result to dist twice (as ref and query). There has been some third-party effort to streamline this process since we haven't gotten to it (see https://github.com/lskatz/mashtree, #9, #66).

[ZeweiSong]

Thanks, I got it like this:

Create sketch for all sequences (into a single file)

mash sketch *.fa -o genomes

Compare individual sequence file with the sketch, and saved into a single dist file

echo *.fa | xargs -n 1 mash dist genomes.msh > dist.txt

I then just parser the txt file in python to keep the non-redundant pairs.

[alienzj]

For the second step:
mash dist genomes.msh genomes.msh > dist.txt