clark stuck during db creation

[Andrea Spitaleri]

Hi all

I am installing clark on our cluster and since 2 days it the db creation.

The command I am running is:

./set_targets.sh /beegfs/datasets/buffer/clarks_db bacteria viruses fungi human

After downloaded all fasta, I do not see any file updated in clarks_db directory. The job (via slurm) seems stuck in this step:

/idle/SOFTW/CLARKSCV1.2.6.1/exe/getfilesToTaxNodes /beegfs/datasets/buffer/clarks_db/taxonomy/nodes.dmp /beegfs/datasets/buffer/clarks_db/.bacteria.fileToAccssnTaxID

the .bacteria.fileToAccssnTaxID has been created but nothing is happening.

Any help?

Thanks

[Rachid]

Dear user,

Is this execution of set_targets.sh done immediately after you installed CLARK? Or have you been using CLARK for quite some time?

I'd update the taxonomy information (with updateTaxonomy.sh cf. README file) just in case and retry again.

Please let us know what happens,

Thank you.

Best,

Rachid

[Andrea Spitaleri]

Yes I did immediately after the installation. So I will kill it (slurm process) and restart again. Do I need to delete the files downloaded or set_targets.sh will restart from where it left?

Best,

Andrea

[Rachid]

Hello,

Based on the information you've shared so far:

I recommend suggestion 1: You start fresh on a different working environment/directory. Then install CLARK and follow the standard instructions on the README file (i.e., set_targets, then classify_metagenome).

I did test this and got no problem.

If you want to use your existing environment/dir then suggestion 2: use "clean.sh" then "updateTaxonomy.sh" then retry with set_targets. If this does not work, use suggestion 1 above.

[Andrea Spitaleri]

Hi

I tried 2. but it did not work. 1. yes (clean installation and etc ...)

thanks!