Error in the long read processing

[EmmettPeng]

Hi OPERA-MS makers,

I have tried thet tool using the command

(base) pengxi@ubuntu_r740:~/softwares/OPERA-MS$ perl OPERA-MS.pl --contig-file /home/wangdanrui/HLa/metaspades/HLa_contigs.fasta --short-read1 /home/pengxi/coldseep_grouped/deduplication/HLa_dedupe_R1.fq --short-read2 /home/pengxi/coldseep_grouped/deduplication/HLa_dedupe_R2.fq --long-read /home/pengxi/2021nanopore_data/group/HL_nanopore.fastq --out-dir /home/pengxi/2021nanopore_data/opera_ms_assemble/HLa --polishing --num-processors 24 --no-ref-clustering --no-strain-clustering

and it reported the error with:

Error in the long read processing. Please see log for details /home/pengxi/2021nanopore_data/opera_ms_assemble/HLa/intermediate_files/read_mapping/OPERA-long-read.out /home/pengxi/2021nanopore_data/opera_ms_assemble/HLa/intermediate_files/read_mapping/OPERA-long-read.err.

In the .out file it has:

(base) pengxi@ubuntu_r740:~/softwares/OPERA-MS$ cat /home/pengxi/2021nanopore_data/opera_ms_assemble/HLa/intermediate_files/read_mapping/OPERA-long-read.out
Mapping long-reads using blasr...
***  Elapsed time: 0
Sorting mapping results...
***  Elapsed time: 0
Analyzing sorted results...

And in the .err file it has:

(base) pengxi@ubuntu_r740:~/softwares/OPERA-MS$ cat /home/pengxi/2021nanopore_data/opera_ms_assemble/HLa/intermediate_files/read_mapping/OPERA-long-read.err
/home/pengxi/softwares/OPERA-MS//tools_opera_ms/blasr  -nproc 24 -m 1 -minMatch 5 -bestn 10 -noSplitSubreads -advanceExactMatches 1 -nCandidates 1 -maxAnchorsPerPosition 1 -sdpTupleSize 7 /home/pengxi/2021nanopore_data/group/HL_nanopore.fastq /home/pengxi/2021nanopore_data/opera_ms_assemble/HLa/intermediate_files/user_assembly/contigs.fasta | cut -d ' ' -f1-12 | sed 's/ /\t/g' > opera.map 2> blasr.err
[INFO] 2021-03-08T03:05:45 [blasr] started.
sort -k1,1 -k10,10g  opera.map > opera.map.sort
Argument "not" isn't numeric in numeric lt (<) at /home/pengxi/softwares/OPERA-MS//OPERA-LG/bin//OPERA-long-read.pl line 700, <MAP> line 1.
Use of uninitialized value $data[8] in join or string at /home/pengxi/softwares/OPERA-MS//OPERA-LG/bin//OPERA-long-read.pl line 701, <MAP> line 1.
Use of uninitialized value $data[9] in join or string at /home/pengxi/softwares/OPERA-MS//OPERA-LG/bin//OPERA-long-read.pl line 701, <MAP> line 1.
Argument "supported." isn't numeric in subtraction (-) at /home/pengxi/softwares/OPERA-MS//OPERA-LG/bin//OPERA-long-read.pl line 615, <FILE> line 1.
Argument "" isn't numeric in subtraction (-) at /home/pengxi/softwares/OPERA-MS//OPERA-LG/bin//OPERA-long-read.pl line 615, <FILE> line 1.
Argument "not" isn't numeric in division (/) at /home/pengxi/softwares/OPERA-MS//OPERA-LG/bin//OPERA-long-read.pl line 626.
Illegal division by zero at /home/pengxi/softwares/OPERA-MS//OPERA-LG/bin//OPERA-long-read.pl line 626.

I am not quite familiar with perl language so it confused me a lot. I have tried the test files and the long read process went well. May you explain this error and give a possible solution? Thank you!

Emmett

[jsgounot]

Hello,

thanks for using OPERA-MS ! Could you please share the content of this file : intermediate_files/read_mapping/opera.map.sort.status ?

Regards,
JS

[EmmettPeng]

Hello,

thanks for using OPERA-MS ! Could you please share the content of this file : intermediate_files/read_mapping/opera.map.sort.status ?

Regards,
JS

Thanks for your prompt reply! The content of the file you required is shown below:

(base) pengxi@ubuntu_r740:~/2021nanopore_data/opera_ms_assemble/HLa/intermediate_files/read_mapping$ cat opera.map.sort.status 
ERROR!	Reading	fasta	files	4Gbytes	is	not	supported.			greater | small-contig

My long read FASTQ files are large indeed. The largest file size exceeds 23Gb. Is OPERA-MS capable of handling large files like this?

Your help is appreciated again!

Best regards,
Emmett Peng

[jsgounot]

Looks like it might be your reference sequence which is too large (see this). How big is your reference sequence ? Maybe using an other LR mapper (option --long-read-mapper minimap2) can be a good workaround until a better solution is found.

[EmmettPeng]

Using minimap2 works. Blasr does not support reference file larger than 4Gb.

[jsgounot]

Indeed, a clean solution would be to add pbmm2, however I don't think this will be done soon. I hope that minimap2 results will be good enough for you. Out of curiosity, on which organism are you working with to have a reference size this big ?

[EmmettPeng]

Indeed, a clean solution would be to add pbmm2, however I don't think this will be done soon. I hope that minimap2 results will be good enough for you. Out of curiosity, on which organism are you working with to have a reference size this big ?

I conducted Illumina Novaseq sequencing and Oxford Nanopore sequencing on environmental microbiome samples and de novo assembly, not just one single organism.