[morecore] insufficient memory

[mictadlo]

Hi,
The Contigs.txt and NbV1ChF.fasta are 2.8 G and 2.6G. Minimap2 seems to run out of memory on 2TB machine.

$ ragoo.py -t 8 -g 100 -s -b -gff augustus.hints_utr.gff3 Contigs.txt NbV1ChF.fasta
Mon Oct 28 01:45:52 2019 --- Running : minimap2 -k19 -w19 -t8 ../NbV1ChF.fasta ../Contigs.txt > contigs_against_ref.paf 2> contigs_against_ref.paf.log
Mon Oct 28 01:49:23 2019 --- Reading alignments
Mon Oct 28 01:52:45 2019 --- Getting gff features
Mon Oct 28 01:53:07 2019 --- Getting contigs
Mon Oct 28 01:53:25 2019 --- Finding interchromosomally chimeric contigs
Mon Oct 28 01:53:25 2019 --- Finding break points and breaking interchromosomally chimeric contigs
Mon Oct 28 01:53:45 2019 --- Running : minimap2 -k19 -w19 -t8 ../../NbV1ChF.fasta Contigs.inter.chimera.broken.fa > inter_contigs_against_ref.paf 2> inter_contigs_against_ref.paf.log
Mon Oct 28 01:57:22 2019 --- Reading interchromosomal chimera broken alignments
Mon Oct 28 02:00:58 2019 --- Finding intrachromosomally chimeric contigs
Mon Oct 28 02:01:52 2019 --- Running : minimap2 -k19 -w19 -t8 ../../NbV1ChF.fasta Contigs.intra.chimera.broken.fa > intra_contigs_against_ref.paf 2> intra_contigs_against_ref.paf.log
Mon Oct 28 02:05:25 2019 --- Reading intrachromosomal chimera broken alignments
Mon Oct 28 02:09:19 2019 --- The total number of interchromasomally chimeric contigs broken is 0
Mon Oct 28 02:09:19 2019 --- The total number of intrachromasomally chimeric contigs broken is 6
Mon Oct 28 02:09:19 2019 --- Assigning contigs
Mon Oct 28 02:09:40 2019 --- Ordering and orienting contigs
Mon Oct 28 02:11:01 2019 --- Creating pseudomolecules
Mon Oct 28 05:26:02 2019 --- Aligning pseudomolecules to reference
Mon Oct 28 05:26:02 2019 --- Running : minimap2 -ax asm5 --cs -t8 ../../NbV1ChF.fasta ../ragoo.fasta > pm_against_ref.sam 2> pm_contigs_against_ref.sam.log
/bin/bash: line 1: 10389 Aborted                 (core dumped) minimap2 -ax asm5 --cs -t8 ../../NbV1ChF.fasta ../ragoo.fasta > pm_against_ref.sam 2> pm_contigs_against_ref.sam.log
Traceback (most recent call last):
  File "/home/ubuntu/miniconda3/envs/ragoo/bin/ragoo.py", line 4, in <module>
    __import__('pkg_resources').run_script('RaGOO==1.1', 'ragoo.py')
  File "/home/ubuntu/miniconda3/envs/ragoo/lib/python3.7/site-packages/pkg_resources/__init__.py", line 666, in run_script
    self.require(requires)[0].run_script(script_name, ns)
  File "/home/ubuntu/miniconda3/envs/ragoo/lib/python3.7/site-packages/pkg_resources/__init__.py", line 1469, in run_script
    exec(script_code, namespace, namespace)
  File "/home/ubuntu/miniconda3/envs/ragoo/lib/python3.7/site-packages/RaGOO-1.1-py3.7.egg/EGG-INFO/scripts/ragoo.py", line 754, in <module>
  File "/home/ubuntu/miniconda3/envs/ragoo/lib/python3.7/site-packages/RaGOO-1.1-py3.7.egg/EGG-INFO/scripts/ragoo.py", line 439, in align_pms
  File "/home/ubuntu/miniconda3/envs/ragoo/lib/python3.7/site-packages/RaGOO-1.1-py3.7.egg/ragoo_utilities/utilities.py", line 25, in run
RuntimeError: Failed : minimap2 -ax asm5 --cs -t8 ../../NbV1ChF.fasta ../ragoo.fasta > pm_against_ref.sam 2> pm_contigs_against_ref.sam.log

less pm_contigs_against_ref.sam.log 
[M::mm_idx_gen::55.511*1.79] collected minimizers
[M::mm_idx_gen::60.192*2.17] sorted minimizers
[M::main::60.192*2.17] loaded/built the index for 19 target sequence(s)
[M::mm_mapopt_update::64.517*2.10] mid_occ = 342
[M::mm_idx_stat] kmer size: 19; skip: 19; is_hpc: 0; #seq: 19
[M::mm_idx_stat::69.857*2.01] distinct minimizers: 143929053 (83.21% are singletons); average occurrences: 1.928; average spacing: 9.884
[M::worker_pipeline::3024.805*3.40] mapped 5 sequences
[M::worker_pipeline::4921.171*3.81] mapped 6 sequences
[M::worker_pipeline::7761.560*3.75] mapped 4 sequences
[morecore] insufficient memory

How important is the the allignment from Aligning pseudomolecules to reference or can ragoo.fasta been used? How much more memory do you think I would need?

Thank you in advance,

Michal

[malonge]

Hmm that is interesting. Firstly, yes you can still use ragoo.fasta. The SV calling step is independent.

I am not sure why Minimap2 is running out of memory. For debugging, you can just run the same or similar command outside of RaGOO. What are you assembling?

Thanks

[mictadlo]

The reference is a plant genome and to assembled it we used for it PacBio and Hi-C data. On the other hand, the 'Contigs.txt` is a pure 50x Illumina assembly created by SparseAssembler.

Additionally, I will run minimap2 manually.

Michal

[malonge]

At this point, if you would like to call SVs, I suggest you do your SV calling manually. You can either use minimap2/paftools or nucmer/assemblytics. If you use paftools, you might as well just write your minimap2 alignments to a PAF file rather than SAM format. I also have a wiki page about this here though you would have to generate your own alignments.

[mictadlo]

Hi,
I think the problem was that the assembly contains many small contigs which could not have been assigned to the reference. Additionally, I asked Ragoo to put 100 N's between the unmapped contigs which lead to a Chr0 size of 550,412,653 bp.

The solution was to use -C which Ragoo.

Michal

[lucventurini]

@malonge
Seen as well when trying to use RaGOO on a species with a large genome (>10Gbps). The out-of-memory problem also happens, in my experience, at the earlier stage of creating pseudomolecules.

I am having a work at some of this in my fork (https://github.com/lucventurini/RaGOO), when I will be done, may I open a pull request?

[malonge]

@mictadlo glad to hear you have resolved it. @lucventurini absolutely thanks so much for contributing.

[cmonat]

Hi,

I'm wondering if the pull request as been done and so if it should work without problem on large genomes?
Thank you and have a great day
Cheers

C.

[malonge]

Hi there,

I am currently working on v2, which uses pysam to dramatically reduce the memory requirements. In fact, the memory for small and large genomes should be roughly the same.

I am hoping to come out with v2 in the next month or so. I will reopen the issue so that I can send a note when the new version is ready.

Thanks

[malonge]

Hi there,

RagTag, the successor to RaGOO, is now available here:

https://github.com/malonge/RagTag

RagTag now uses pysam to query read coverage, so the memory requirement is dramatically reduced.

Thanks,
Mike