multithreading smalt error

[lmoncla]

When running IVA on full genome influenza samples and enabling multiple threads, I get the following smalt error:

smalt.c:807 ERROR: The two FASTA/FASTQ input file have different numbers of reads
[W::sam_read1] parse error at line 812880
[main_samview] truncated file.

I do not get this error if I run the same sample with a single thread.

[martinghunt]

Could you share the input reads with me please so I can try to debug? Thanks.

[lmoncla]

Hi, Here they are! They have already been trimmed with trimmomatic. Thanks so much. Best, ​ fastqs.zip <https://drive.google.com/file/d/0B0s7NWreQ5QqWUpNcnFxcVB4LVE/view?usp=drive_web> ​ Louise

…

On Tue, Feb 14, 2017 at 5:14 AM, martinghunt ***@***.***> wrote: Could you share the input reads with me please so I can try to debug? Thanks. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#71 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AIgbALLC_dMnKgeAyq6S1OIvqE-6Ier_ks5rcYyegaJpZM4L_nFS> .

[martinghunt]

Thanks. What version of IVA did you use and what was the command to run it?

[lmoncla]

IVA version 1.0.8 Using kmc version 2.1.1 Using kmc_dump version 2.1.1 Using nucmer version 3.1 Using samtools version 1.3.1 Using smalt version 0.7.6 iva -f 6390_S23_L001_R1_001.trimmed.fastq -r 6390_S23_L001_R2_001.trimmed.fastq IVA_output

…

On Tue, Feb 14, 2017 at 10:12 AM, martinghunt ***@***.***> wrote: Thanks. What version of IVA did you use and what was the command to run it? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#71 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AIgbAACWUktxVjd5olu71k0hjMtu5cYZks5rcdJcgaJpZM4L_nFS> .

[martinghunt]

I don't yet know why you didn't get the error with a single thread, but the reason for the error is that there needs to be the same number of reads in the two files. And the N^th read in each file should be mate pairs.

$ wc -l *
  1989088 6390_S23_L001_R1_001.trimmed.fastq
  1657760 6390_S23_L001_R2_001.trimmed.fastq

If you trim with trimmomatic, then you'll need to only keep the paired reads. Throw away everything else, ie when one read of a pair is removed by trimmomatic.

[lmoncla]

Thank you, I will try that. I assume that it is not possible to run IVA on unpaired reads?

…

On Tue, Feb 14, 2017 at 10:40 AM, martinghunt ***@***.***> wrote: I don't yet know why you didn't get the error with a single thread, but the reason for the error is that there needs to be the same number of reads in the two files. And the N^th read in each file should be mate pairs. $ wc -l * 1989088 6390_S23_L001_R1_001.trimmed.fastq 1657760 6390_S23_L001_R2_001.trimmed.fastq If you trim with trimmomatic, then you'll need to only keep the paired reads. Throw away everything else, ie when one read of a pair is removed by trimmomatic. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#71 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AIgbAHivyMEqmsuJV2c87yoNM3H3XJR8ks5rcdkXgaJpZM4L_nFS> .

[martinghunt]

No, sorry but the read pair information is essential to its algorithm.

[lmoncla]

I see, I thought so. Thanks for the very quick reply, I really appreciate it!

…

On Tue, Feb 14, 2017 at 10:44 AM, martinghunt ***@***.***> wrote: No, sorry but the read pair information is essential to its algorithm. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#71 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AIgbAANS_3Wf0VQ-TdBx_RnZBUBo_lMBks5rcdn-gaJpZM4L_nFS> .