denovo_map.pl error: "Error: Unable to open tag file for writing"

[Kimberly Neil]

I'm running stacks 2.0 on my university's cluster with the following script:

denovo_map.pl -M 3 -n 0 -T 15 -o /path/to/output/directory --popmap /path/to/map --samples /path/to/samples -X "populations: -p 3 -r 0.80 --fstats --genepop --vcf"

I'm doing this on 272 samples that have already gone through the process rad command. I checked the log, and it goes through sample by sample as expected. When it gets to sample 231, it reads the infput file, loads the rad stacks, estimates stack coverage, assembled the stacks, merges secondary stacks, estimates final coverage, and then fails at this step:

"Calling consensus sequences and haplotypes for catalog assembly...

 Writing tags, SNPs, and alleles files...

 Refetching read IDs...Error: Unable to open tag file for writing.

denovo_map.pl: Aborted because the last command failed (1)."

I'm going to try deleting that sample and continuing. But do you know what went wrong? 

Thanks!

[Julian Catchen]

I would check that you didn't run out of disk space.

Kimberly Neil wrote on 8/6/18 10:18 AM:

[Kimberly Neil]

Thanks! 

I think you're right - I had exceeded the disk space allocated to my cluster account. I realized the original files I had weren't zipped, so I did that to save space. Do I need to alter my script in any way to indicate the files are .fq.gz now? I submitted the job again, and the log has been on "Sample 1 of 272" for a while now. Does stacks take longer when the files are .gz vs. .fq (when I initially used the .fq files, ustacks was pretty fast per sample)?

Thanks very much.