Xargs: cat: terminated by signal 13 with kraken2-build --build.

[kdbchau]

Hello,

I am using kraken2 version 2.1.2 and I am trying to build the nt database with the following command:

kraken2-build --build --db /scratch/chauk/kraken2/kraken_nt/nt --threads 48

It fails at step 2:

Creating sequence ID to taxonomy ID map (step 1)...
Found 81034768/81152031 targets, searched through 797981066 accession IDs, search complete.
lookup_accession_numbers: 117263/81152031 accession numbers remain unmapped, see unmapped.txt in DB directory
Sequence ID to taxonomy ID map complete. [1h12m58.303s]
Estimating required capacity (step 2)...
xargs: cat: terminated by signal 13

I tried again with changing the threads to 1:

Creating sequence ID to taxonomy ID map (step 1)...
Sequence ID to taxonomy ID map already present, skipping map creation.
Estimating required capacity (step 2)...
xargs: cat: terminated by signal 13

What could be the problem?

[SaraOliva]

I have the same problem. Could you solve it?

[kdbchau]

Nope never solved. I just stopped bothering with kraken2. Seems too buggy and no one is helping.

[mclaugsf]

i'm guessing you're running out of memory. I just submitted this ticket: https://github.com/jenniferlu717/Bracken/issues/236

i was really having trouble with plant genomes which are low complexity and highly enriched for repeats. I think building a database of plant genomes results in finding kmers that are very frequent which can use quite a bit of memory and I also think there's a memory leak in kraken2 so if you keep running it on larger sets of plant genomes it will use much more memory than you think it should. i solved the issue by chopping it up into smaller chunks which confirms there's a memory leak issue.

[marco315176]

I downloaded the taxonomi directory with:

kraken2-build --download-taxonomy --db ${db}

and it was solved