Thank you @Zer0day-0 .

I went back and looked at my notes on the last time I tried to install with Conda: #736 (comment)

I noticed that you're using CentOS. I think I might be testing with CentOS at the time. So let me try that.

Get a CentOS machine to test.

I used:

gcloud compute instances create "${USER}-centos7" \
--scopes "compute-rw,storage-full,cloud-platform" \
--image-family "centos-7" \
--image-project "centos-cloud" \
--machine-type "e2-standard-16" \
--zone "us-west1-b"

gcloud compute ssh pichuan-centos7 --zone "us-west1-b"

Version:

$ uname -a
Linux pichuan-centos7 3.10.0-1160.114.2.el7.x86_64 #1 SMP Wed Mar 20 15:54:52 UTC 2024 x86_64 x86_64 x86_64 GNU/Linux

Install conda:

curl https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh > Miniconda3-latest-Linux-x86_64.sh
bash Miniconda3-latest-Linux-x86_64.sh -b -u -p $HOME/miniconda
eval "$(${HOME}/miniconda/bin/conda shell.bash hook)"

To repeat what I did in, I tried: #736 (comment)

conda config --add channels defaults && \
conda config --add channels bioconda && \
conda config --add channels conda-forge
conda create -y -n dv-env deepvariant
conda activate dv-env

It completed without any error messages. I see:

(dv-env) [pichuan@pichuan-centos7 ~]$ ls /home/pichuan/miniconda/envs/dv-env/share/
bash-completion  deepvariant-1.5.0-0  doc  et  examples  google-cloud-sdk-359.0.0-0  icu  info  keyutils  licenses  locale  man  tabset  terminfo  zsh
(dv-env) [pichuan@pichuan-centos7 ~]$ ls /home/pichuan/miniconda/envs/dv-env/share/deepvariant-1.5.0-0/binaries/DeepVariant/1.5.0/DeepVariant-1.5.0
call_variants_keras.zip  freeze_graph.zip           make_examples.zip  multisample_make_examples.zip  runtime_by_region_vis.zip  vcf_stats_report.zip
call_variants.zip        licenses.zip               model_eval.zip     postprocess_variants.zip       settings.sh
deeptrio                 make_examples_somatic.zip  model_train.zip    run-prereq.sh                  show_examples.zip

As mentioned in my previous investigation, I don't actually know how to use Conda though. But it seems like at least with the setting above, it can install.

If you can try that and see if it works for you. And do what you'd usually do to run and see if it works, that will be great! Please let me know.

By the way,
I noticed you were using bioconda/label/cf201901::deepvariant. I actually don't know what that does. (Can you explain where that came from and why you needed that?)
But to understand be behavior, I also tried something like:
conda create -y -n dv-env bioconda/label/cf201901::deepvariant

If I did that, then it still seems to be work, but it somehow had an older version deepvariant-0.7.2-1 in /home/pichuan/miniconda/envs/dv-env/share/, which I don't understand why.

I'd recommend that you run using my first command, so you can have a more recent version to try out!

from deepvariant.

Hi @Zer0day-0 ,
thanks for the report. DeepVariant conda actually isn't maintained by our team. I can take a look later this week.

But before that , I have a question - is it possible for you to use other solutions, such as our Docker version or Singularity?
If you can describe your situation and why you might only be able to use conda, that will also be useful context for us.

Thank you!

from deepvariant.

Hello @pichuan,
Thanks for response, actually I am running a SGE cluster in order to run my analysis, and the sysadmins keeps use of docker restricted(root access issue). Hence I am tryng to use conda.

from deepvariant.

Thank you. Would you be able to use Singularity?

from deepvariant.

Nope, similar issue of sysadmin!

from deepvariant.

thank you for your kind support @pichuan

It seems I was running a really old and obsolete version of anaconda(4.2.1) in my SGE. It took some gymnastics to install it for a single user and run across the nodes but I updated my conda to current version. after that installation completed with no errors.

the "bioconda/label/cf201901::deepvariant" was provided in bioconda page, which is supposed to be another installation solution. I think this issue can be closed. For anyone in future running in a problem similar to me, I'll make a repo with the installation code in my profile!

from deepvariant.