Jellyfish problem with Failed to open input file 'reads.jf'

Question

[m.t.lorenc]

Hi, I can't get jellyfish running with two different ways:

Method 1

jellyfish count -t 10 -C -m 21 -s 10G --quality-start=33 -o reads.jf <(zcat V350181330_L03_R1.fastp.fq.gz) <(zcat V350181330_L04_R1.fastp.fq.gz)

26897 Killed                  jellyfish count -t 10 -C -m 21 -s 10G --quality-start=33 <(zcat V350181330_L03_R1.fastp.fq.gz) <(zcat V350181330_L04_R1.fastp.fq.gz) -o reads.jf
Failed to open input file 'reads.jf'

Method 2

zcat V*_R1.fastp.fq.gz | jellyfish count -t 10 -C -m 21 -s 10G --quality-start=33 -o reads.jf

zcat V*_R1.fastp.fq.gz
     53359 Killed                  | jellyfish count -t 10 -C -m 21 -s 10G --quality-start=33 -o reads.jf
Failed to open input file 'reads.jf'

What did I miss?

[Joe]

Where are you running the command? Do you/the tool have write privileges in the folder?

[Brian Bushnell]

Often when I get mysterious 'killed' messages it's because the OS had some limit of how much memory I could use, and I exceeded it. It might be helpful here to mention the platform specifications, whether it was shared, and what 'top' said about the memory available and what was running. Obviously, the size of the input file(s) is important to disclose as well.

Using BBTools you can get a cardinality estimate like this:

zcat V*_R1.fastp.fq.gz | loglog.sh in=stdin.fq

That will let you know how many unique kmers are present in your data, using a fixed amount of memory. Then you can estimate the amount of memory needed by multiplying that by 12 or however many bytes Jellyfish uses to store kmers.

[m.t.lorenc]

Thank you. I increased the memory, and Jellyfish finished.