Assertion in bntseq fails

[ahwagner]

BWA crashes when running on my PE RNA-seq data.

It looks like the failure is caused by a failed assertion here.

The logged error is as follows:

[M::bwa_idx_load_from_disk] read 0 ALT contigs
[M::process] read 858446 sequences (80000186 bp)...
[E::bns_fetch_seq] begin=109745949, mid=109746080, end=95461170, len=14284779, seq=0x7ff5c225e010, rid=21138, far_beg=95459107, far_end=95461170
bwa: bntseq.c:444: bns_fetch_seq: Assertion `seq && *end - *beg == len' failed.

[blacksxn]

I have the same question when using the latest version of BWA (0.7.15). while when i change the version to 0.7.8, the question disappeared. i really confused about the result.

[nanakiksc]

This seems to happen if you use a named pipe when creating the index, as in:
$ bwa index <(zcat ref.fa.gz)
Try redoing the index, this time from an already decompressed file, until the issue is fixed.
Hope this helps.

PS: It happens both in the latest version from GitHub (0.7.15-r1144-dirty) and in the version from the Canonical repo (0.7.12-r1039).

[ALFLAG]

Hello every one,
I have the same error message after trying to use bwa mem on a sequence of more than 4.8 billions nucleotides.
Indexation was done with the following parameter -b 600000000 and -a bwtsw, on a decompressed fasta file.
My bwa version is 0.7.17-r1188 and was installed with conda.

Any suggestions ?
Thanks !

[yanlinlin82]

@ALFLAG could you please help to try this pull request #243

[ALFLAG]

I'd love to. But I can't find the bntseq.c file. Is it because the tool has been installed with conda ?

[yanlinlin82]

You may try like this, build from the source, and try this test version:

git clone https://github.com/yanlinlin82/bwa.git -b patch-1 bwa-patch-1
make -C bwa-patch-1/
./bwa-patch-1/bwa

[ALFLAG]

@yanlinlin82
I test the patch you've made, no error message, the indexation succeed.
Thanks a lot.

[yanlinlin82]

Thank you for testing and feeding back.

[AliSajid]

Reopening this as I am facing the exact same issue:

Here's the output from my command:
bwa mem -R '@RG\tID:bwa\tSM:A\tPL:ILLUMINA\tPU:HiSEQ2000' -o output_pe.sam ../GRCH38/Homo_sapiens.GRCh38.dna.primary_assembly.fa SRR3438555_1.fastq SRR3438555_2.fastq [M::bwa_idx_load_from_disk] read 0 ALT contigs [M::process] read 200000 sequences (10000000 bp)... [M::process] read 200000 sequences (10000000 bp)... [E::bns_fetch_seq] begin=3324490029, mid=3324490030, end=3099750718, len=224739311, seq=0x2ba8a4000010, rid=193, far_beg=3099749748, far_end=3099750718 bwa: bntseq.c:450: bns_fetch_seq: Assertion seq && *end - *beg == len' failed.
Aborted`

I tried to do understand the error. It complained about an equality. Here's the output from bc:

3099750718 - 3324490029 -224739311

The length is the same except it has a negative sign.

Edit:

I tried to run the patch. Here's the output from the patch:

/users/PJS0313/asimami/scratch/bwa-patch-1/bwa mem -R '@RG\tID:bwa\tSM:A\tPL:ILLUMINA\tPU:HiSEQ2000' -o output_pe.sam ../GRCH38/Homo_sapiens.GRCh38.dna.primary_assembly.fa SRR3438555_2.fastq SRR3438555_1.fastq [M::bwa_idx_load_from_disk] read 0 ALT contigs [M::process] read 200000 sequences (10000000 bp)... [M::process] read 200000 sequences (10000000 bp)... bwa: bwamem.c:750: mem_chain2aln: Assertion re >= 0' failed.
Aborted`

[ZZH19999]

BWA在运行我的PE RNA-seq数据时崩溃。

看起来失败是由此处的失败断言引起的。

记录的错误如下所示：

[M::bwa_idx_load_from_disk] read 0 ALT contigs
[M::process] read 858446 sequences (80000186 bp)...
[E::bns_fetch_seq] begin=109745949, mid=109746080, end=95461170, len=14284779, seq=0x7ff5c225e010, rid=21138, far_beg=95459107, far_end=95461170
bwa: bntseq.c:444: bns_fetch_seq: Assertion `seq && *end - *beg == len' failed.

How did you solve this problem

[hfpef]

I run into the same problem. I tried to increase memory at first, but it still didn't work and I solved it by using bwa-mem2.

First I installed bwa-mem2:

curl -L https://github.com/bwa-mem2/bwa-mem2/releases/download/v2.0pre2/bwa-mem2-2.0pre2_x64-linux.tar.bz2 \
  | tar jxf -

and instead of using the standard:

bwa mem ref.fa read1.fq read2.fq > aln-pe.sam

after installing bwa-mem2 the code will look like this:

bwa-mem2-2.0pre2_x64-linux/bwa-mem2 mem ref.fa read1.fq read2.fq > aln-pe.sam

In my case it looked like this:

bwa-mem2-2.0pre2_x64-linux/bwa-mem2 mem -R "@RG\tID:id\tSM:nr\tLB:nr" chr1.fa chr1_ptn_R1.fastq chr1_ptn_R2.fastq \
| samblaster --excludeDups --addMateTags --maxSplitCount 2 --minNonOverlap 20 \
| samtools view -S -b - \
 > sample_ptn_chr1.bam

I hope it's still useful and it will work for you guys!

PS I will try to give an update to make bwa-mem2 possible to be used while running speedseq or other tools that base on bwa algorythm.

[heihaheihaha]

Reindex the reference fasta helps me under the same problem.