Resfinder download fails

smdabdoub commented 3 years ago

Expected Behavior

The Resfinder database is downloaded and an mmseqs2 index is created.

Current Behavior

Fails at the download extraction step.

Steps to Reproduce (for bugs)

mmseqs databases Resfinder db/rf tmp

MMseqs Output (for bugs)

databases Resfinder db/rf tmp

MMseqs Version: 13.45111 Force restart with latest tmp false Remove temporary files false Compressed 0 Threads 48 Verbosity 3

--2021-10-08 14:50:24-- https://api.bitbucket.org/2.0/repositories/genomicepidemiology/resfinder_db/commit/master?fields=hash,date Resolving api.bitbucket.org (api.bitbucket.org)... 18.205.93.8, 18.205.93.6, 18.205.93.7, ... Connecting to api.bitbucket.org (api.bitbucket.org)|18.205.93.8|:443... connected. HTTP request sent, awaiting response... 200 OK Length: 89 [application/json] Saving to: 'tmp/9391455786971071082/version'

tmp/9391455786971071082/version 100%[=================================================================================================>] 89 --.-KB/s in 0s

2021-10-08 14:50:24 (13.7 MB/s) - 'tmp/9391455786971071082/version' saved [89/89]

--2021-10-08 14:50:24-- https://bitbucket.org/genomicepidemiology/resfinder_db/get/master.tar.gz Resolving bitbucket.org (bitbucket.org)... 104.192.141.1, 2406:da00:ff00::22c3:9b0a, 2406:da00:ff00::34cc:ea4a, ... Connecting to bitbucket.org (bitbucket.org)|104.192.141.1|:443... connected. HTTP request sent, awaiting response... 200 OK Length: 457328 (447K) [application/x-tar-gz] Saving to: 'tmp/9391455786971071082/master.tar.gz'

tmp/9391455786971071082/master.tar.gz 100%[=================================================================================================>] 446.61K --.-KB/s in 0.07s

2021-10-08 14:50:25 (5.88 MB/s) - 'tmp/9391455786971071082/master.tar.gz' saved [457328/457328]

tar: .fsa: not found in archive createdb tmp/9391455786971071082/.fsa db/rf --compressed 0 -v 3

Input tmp/9391455786971071082/*.fsa does not exist Error: createdb died

Context

Your Environment

Include as many relevant details about the environment you experienced the bug in.

Git commit used (The string after "MMseqs Version:" when you execute MMseqs without any parameters): h95f258a
Which MMseqs version was used (Statically-compiled, self-compiled, Homebrew, etc.): singularity container downloaded from Biocontainers
Server specifications (especially CPU support for AVX2/SSE and amount of system memory): Intel Xeon 8268, 192GB RAM
Operating system and version: RHEL 7.9

milot-mirdita commented 3 years ago

This is quite confusing. For whatever reason your system's tar doesn't use the default --wildcards option. It seems to be having --no-wildcards set as the default mode. I don't have access to RHEL, but in the centos 7.9 docker image this doesn't seem to happen.

You can probably get around this issue by doing:

TAR_OPTIONS="--wildcards" mmseqs databases Resfinder db/rf tmp

smdabdoub commented 3 years ago

Ah, ok that actually does make sense. It's not using the OS tar because I was running it through a singularity container. So I'm guessing whatever tar was built into the container had the issue. I tried the standard command again but with a conda install of mmseqs2 and it worked fine. Thanks!

milot-mirdita commented 3 years ago

Is this a public container? I would like to know how exactly this failed to improve the script for the future.

smdabdoub commented 3 years ago

It is. The container was created by the BioContainers people. Although I can never get their registry to load. So I downloaded it from quay.io using this link: https://quay.io/repository/biocontainers/mmseqs2?tab=tags

soedinglab / MMseqs2