bwa-mem2:2.2.1--he513fc3_0 Container Missing Executable

[pnolabs]

I am trying to use the container quay.io/biocontainers/bwa-mem2:2.2.1--he513fc3_0, but bwa-mem2 does not seem to be available in the expected location. Running the following command:
docker run --rm quay.io/biocontainers/bwa-mem2:2.2.1--he513fc3_0 bwa-mem2 version
Results in:
ERROR: fail to find the right executable
However, checking inside the container shows that the executable exists at /usr/local/bin/bwa-mem2.

Can you confirm if this is a packaging issue?

Originally, I came across this error when trying to run the nf-core/raredisease pipeline with nextflow.

Docker version 27.5.1, build 9f9e405
nextflow version 24.10.3.5933
conda 24.7.1
nf-core 3.1.1 pypi_0 pypi
python --version Python 3.9.20
nf-core/raredisease v2.2.0-gfa61a65

[TaskFinalizer-2] ERROR nextflow.processor.TaskProcessor - Error executing process > 'NFCORE_RAREDISEASE:RAREDISEASE:PREPARE_REFERENCES:BWAMEM2_INDEX_GENOME (reference.fasta)'

Caused by:
Process NFCORE_RAREDISEASE:RAREDISEASE:PREPARE_REFERENCES:BWAMEM2_INDEX_GENOME (reference.fasta) terminated with an error exit status (2)
...
Command error:
Unable to find image 'quay.io/biocontainers/bwa-mem2:2.2.1--he513fc3_0' locally
2.2.1--he513fc3_0: Pulling from biocontainers/bwa-mem2
1dbcab28ce46: Pulling fs layer
cfb1ba34637d: Pulling fs layer
ace2d8a63dd5: Pulling fs layer
75c080ef15eb: Pulling fs layer
1dbcab28ce46: Waiting
cfb1ba34637d: Waiting
...
2e178fd72baf: Pull complete
b28b1e77606a: Pull complete
Digest: sha256:8489086047b16f49a01a077065c88258c28027c49c35281f77df353bcc370970
Status: Downloaded newer image for quay.io/biocontainers/bwa-mem2:2.2.1--he513fc3_0
ERROR: fail to find the right executable

Thank you very much for your time and help!

[mboudet]

Hello!

All images available on quay.io are built from conda packages, defined in the https://github.com/bioconda/bioconda-recipes repository.

So you may need to make an issue on that side.

However, this is weird, as I cannot reproduce the issue.. on my end:

docker run --rm quay.io/biocontainers/bwa-mem2:2.2.1--he513fc3_0 bwa-mem2 version
nable to find image 'quay.io/biocontainers/bwa-mem2:2.2.1--he513fc3_0' locally
2.2.1--he513fc3_0: Pulling from biocontainers/bwa-mem2
1dbcab28ce46: Pull complete 
cfb1ba34637d: Pull complete 
ace2d8a63dd5: Pull complete 
75c080ef15eb: Pull complete 
316957f8baaf: Pull complete 
dbd31e1d863d: Pull complete 
2f8531d5a6ec: Pull complete 
2e178fd72baf: Pull complete 
b28b1e77606a: Pull complete 
Digest: sha256:8489086047b16f49a01a077065c88258c28027c49c35281f77df353bcc370970
Status: Downloaded newer image for quay.io/biocontainers/bwa-mem2:2.2.1--he513fc3_0
Looking to launch executable "/usr/local/bin/bwa-mem2.avx512bw", simd = .avx512bw
Launching executable "/usr/local/bin/bwa-mem2.avx512bw"
2.2.1

Seems like there is some weird $PATH issues.. but even with the same docker version, I cannot reproduce it.

[pnolabs]

@mboudet thank you for your quick reply!

I came across another post that points to a hardware issue:
bwa-mem2/bwa-mem2#73

I have a Vendor ID: GenuineIntel, Model name: Common KVM processor, CPU(s): 16
and for me this command gives no output:
egrep --color 'sse4_1|avx2' /proc/cpuinfo

Does this mean my processor does not support SSE4.1. which seems to be required for running bwa-mem2?

[mboudet]

Ah, it makes sense. I read the ERROR: fail to find the right executable as a command not found, but it's thrown by bwa-mem itself, who tries to find the correct executable for your hardware, and fails.

On my end, egrep --color 'sse4_1|avx2' /proc/cpuinfo works.

This is just a guess, but maybe you are running on a VM? The 'model name : Common KVM processor' does seem to imply you are using an virtualized CPU (and not the host one), and bwa-mem does not know what to do with it.

In any case, this is not a packaging issue, but a hardware compatibility one.

[pnolabs]

@mboudet yes, I am running it on a VM. Sorry for missing to edit this information.

$ lscpu
Architecture: x86_64
CPU op-mode(s): 32-bit, 64-bit
Address sizes: 40 bits physical, 48 bits virtual
Byte Order: Little Endian
CPU(s): 16
On-line CPU(s) list: 0-15
Vendor ID: GenuineIntel
Model name: Common KVM processor
CPU family: 15
Model: 6
Thread(s) per core: 1
Core(s) per socket: 16
Socket(s): 1
Stepping: 1
BogoMIPS: 4399.99
Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall
nx lm constant_tsc nopl xtopology cpuid tsc_known_freq pni cx16 x2apic hypervisor lahf_lm cpuid_fault pti s
sbd ibrs ibpb

So, I guess this means i cannot run bwa-mem2 on my VM with the given hardware? And should switch to older softwares like bwa?

[mboudet]

I think you can change some stuff in your virtualisation software to use your host cpu, and not the emulated one. (ex: https://forum.proxmox.com/threads/common-kvm-processor.66160/)

But this is software-dependant, so I'm not really able to help much further.

[pnolabs]

@mboudet Thank you for helping and pointing out a possible solution.