Profile selection in conda environment

[burcudsr]

Dear Poseidon team,

I have a question about -profile selection. I am trying to run example file called "bats_mx1_small.fasta" on a HPC using Singularity but also running nextflow via miniconda environment.

When i set -profile slurm,conda give me an error like:

Unknown configuration profile: 'conda'
Did you mean one of these?
yoda

When i set -profile slurm,singularity give me an error like:

Error executing process > 'check_fasta_format (1)'
Caused by:
Failed to submit process to grid scheduler for execution
Command executed:
sbatch .command.run
Command exit status:
1
Command output:
sbatch: error: Batch job submission failed: Unspecified error

Do you have any suggestions about that, i am new to nextflow. I will be glad for your contributions.

[hoelzer]

Dear @burcudsr , thanks for using PoSeiDon!

Unfortunately, there is currently no conda env available. The pipeline is tested with Docker support which might be only available on your HPC via Singularity. However, I did not fully test the pipeline under singularity yet.

You can give it a try via -profile slurm,singularity

I hope that works, otherwise please just let me know!

[burcudsr]

Dear @hoelzer , thanks for your quick reply.

I tried with -profile slurm,singularity on HPC and can't find singularity command which is in /usr/bin/singularity . However, I managed to run test_file -profile local,docker without any errors on my local computer, so I'll try to run my files from there.

[hoelzer]

Hi,

on your HPC you maybe have to load singularity as a module:

module load singularity

is

singularity --version

working?

Otherwise, when your data is not to large it should also run w/ Docker on a laptop (just might need some more time)

[Victory-Hugo]

Dear Developer,

I hope this message finds you well. I would like to express my sincere appreciation for your excellent software, poseidon. It has provided a wonderfully integrated workflow for investigating whether genes are under positive selection, and it has been tremendously helpful in my research.

I have two humble suggestions that I hope may further improve the user experience.

First, in practical use, high-performance computing servers often have unstable network connections, and the Docker image of poseidon is relatively large. I noticed that your software allows specifying an offline conda environment, which is very helpful. However, for users who are less familiar with configuration and deployment, the instructions for this feature may not be entirely intuitive. Clearer guidance or examples might be beneficial.

Second, it would be great if poseidon could allow users to customize which analysis steps to run. Some users may only need to perform the first one or two steps rather than the full pipeline. If each step could be enabled or disabled through a simple setting, it would make the software even more flexible and user-friendly.

Thank you very much for your time and for developing such a valuable tool. I truly appreciate your work and look forward to future updates of poseidon.

[hoelzer]

Dear @Victory-Hugo thanks for the kind words and interest in the tool!

I agree that the documentation could be improved regarding Conda usage. On the other hand, PoSeiDon uses multiple Docker containers for the different steps so I think downloading them on an HPC (you probably would like to use Singularity then, not Docker) should be fine.

Customization of the analyses steps should be also fairly simple with some parameters.

However, pleae note that I am only able to maintain PoSeiDon in my free time which will be especially limited the next months. I keep all issues open and eventually can manage to get back to them.

Pull Requests adding new features are also always welcome : )

cheers