bcgsc / abyss

:microscope: Assemble large genomes using short reads
http://www.bcgsc.ca/platform/bioinfo/software/abyss
Other
308 stars 108 forks source link

Can not run the assembly on a system with 128GB RAM #209

Closed aidashah closed 6 years ago

aidashah commented 6 years ago

I have a whole genome sequencing dataset of an insect with a genome size around 500Mb. I have tried to assemble it via Abyss but, it is killed, most probably due to RAM insufficiency. I have asked the developers and they recommended using another mode of the program. Actually, apparently, the program has two modes, one is hash table and another one is Bloom filter. The first one uses so much RAM capacity (around 200GB for a genome size of 500Mb) while the other one needs much less RAM. I have run the bloom filter mode according to ( https://github.com/bcgsc/abyss#assembling-using-a-bloom-filter-de-bruijn-graph) and it is alright for now (It is not finished yet). I am sharing this here, in case that others might encounter the same trouble in the future.

sjackman commented 6 years ago

Glad to hear that you got it working, and thanks for reporting your experience.

aidashah commented 6 years ago

Dear Shuan, My run has been finished withe following messages: Minimum and maximum distance are set to -63 and 971 bp. error: no @SQ records in the SAM header make: [tpittrim10643-3.dist] Error 1 make: Deleting file `tpittrim10643-3.dist' As, it is my first experience, I don't know whether it is finished successfully or not.

sjackman commented 6 years ago

The run was unsuccessful. Please report the exact command that you ran and the complete log file of your run.

aidashah commented 6 years ago

The following command I have run: abyss-pe name=tpittrim10643 k=64 in='trimmed10-R1.fastq trimmed10-R2.fastq' B=64G H=3 kc=3 v=-v After the following commands the assembly has been stopped.

FR Stats mean: 398.1 median: 392 sd: 93.62 n: 147696758 min: 58 max: 971 ignored: 19183 ▁▁▂▃▄▅▆▇██████▇▆▆▅▄▃▃▂▂▁▁____ sort: write failed: /tmp/sortlvR4ke: No space left on device Mate orientation FR: 147711567 (100%) RF: 4374 (0.00296%) The library tpittrim10643-3.hist is oriented forward-reverse (FR). Stats mean: 398.1 median: 392 sd: 93.62 n: 147696758 min: 58 max: 971 ▁▁▂▃▄▅▆▇██████▇▆▆▅▄▃▃▂▂▁▁____ Minimum and maximum distance are set to -63 and 971 bp. error: no @SQ records in the SAM header make: [tpittrim10643-3.dist] Error 1 make: Deleting file `tpittrim10643-3.dist'

aidashah commented 6 years ago

No log file is generated.

aidashah commented 6 years ago

Hello, I have written the command under the same issue in the github. I would be so thankful if you can take a look. Regards,

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/bcgsc/abyss/issues/209#issuecomment-377730627, or mute the thread https://github.com/notifications/unsubscribe-auth/AjdXpxeaWFUQ-mrP6x7j9K-GtnANzyRHks5tkBBWgaJpZM4TBn50 .

-- Aida Shahraki, M.Sc. Bogazici University Department of Molecular Biology and Genetics Cancer Signaling Laboratory (NBI_CSL) Kuzey Park building, 34342 Bebek-Istanbul/Turkey

kmnip commented 6 years ago

This is the relevant part of your log:

sort: write failed: /tmp/sortlvR4ke: No space left on device

Your temp directory is full. You may want to redirect the TMPDIR environment variable to your current working directory before running ABySS.

benvvalk commented 6 years ago

thanks @sjackman and @kmnip

sjackman commented 6 years ago

On some systems /var/tmp has more space than /tmp.

df -h /tmp /var/tmp
export TMPDIR=/var/tmp