search

BioInf-Wuerzburg / proovread

PacBio hybrid error correction through iterative short read consensus
MIT License
60 stars 20 forks source link

Can't allocate memory while bwa indexing #154

Closed yazuo0204 closed 4 years ago

yazuo0204 commented 4 years ago

Hi,

I have three Nanopore RNA sequence need to be corrected. But one of them would report Can't allocate memory while running bwa index. I am sure that I have enough memory -- it still has 500G RAM for running this one. And other two sequence running ok without any errors. I tried to change the threads to 1 or default, but it didn't working.

The *.masked.fasta has 9G, so I thought it might due to the sequence is extremely large. I tried to drag out 1000 reads to run proovread-bwa index and it didn't output any errors

` /data/user/zuoya/Biosoft/proovread/bin/proovread -t 1 --long-reads=$filter_reads --short-reads=$fq1 --short-reads=$fq2

------------------------------------------------------------------------------

[Wed Jul 1 16:21:46 2020] Read IDs don't look like proper PacBio subread IDs. Falling back to running mode: sr-noccs

[Wed Jul 1 16:22:16 2020] Running task bwa-sr-1 [Wed Jul 1 16:22:16 2020] Indexing Long reads

------------------------------------------------------------------------------

/data/user/zuoya/Biosoft/proovread/bin/../util/bwa/bwa-proovread index proovrea\ d/read-long/proovread.masked.fa proovread/read-long/proovread.masked.fa 2>proov\ read/bwa-sr-1/proovread_bwa.log

------------------------------------------------------------------------------

------------------------------------------------------------------------------

    Exited at 'main'
    '/data/user/zuoya/Biosoft/proovread/bin/proovread', line 1277
    Last call 'main::run_bwa'
    '/data/user/zuoya/Biosoft/proovread/bin/proovread', line 853

[bwa_index] Pack FASTA... [add1] Failed to allocate -34359738368 bytes at bntseq.c line 260: Cannot allocate memory `

thackl commented 4 years ago

How big is the file with the nanopore reads? I think there is a maximum file size that bwa can index.

Try runnning /path/to/proovread/util/bwa/bwa-proovread index $filter_reads directly in the command line. If that does not work, this is a bwa issue. You will have to split your input file into smaller chunks. Have a look at https://github.com/BioInf-Wuerzburg/proovread#getting-chunk-sizes-right-48