Baysor module #7270
Baysor module #7270
Conversation
modules/nf-core/baysor/main.nf
Outdated
| process BAYSOR { | ||
| tag '$meta.id' | ||
| label 'process_high_memory' | ||
| container "docker.io/segonzal/baysor:0.7.1" |
There was a problem hiding this comment.
@sebgoti, ask a core team member to get the baysor image into nf-core
You can also refer to Docker image if the image size is too large
There was a problem hiding this comment.
@khersameesh24, indeed my current image is very large (~15 GB). I could refer to yours but I wanted to use the latest version (0.7.1). Could you help me figuring out how to reduce the size? Here is how I build the image:
FROM quay.io/pypa/manylinux2014_x86_64
WORKDIR /app
RUN yum install -y unzip
RUN curl -O -L https://github.com/kharchenkolab/Baysor/releases/download/v0.7.1/baysor-x86_x64-linux-v0.7.1_build.zip
RUN unzip baysor-x86_x64-linux-v0.7.1_build.zip
RUN mv bin unzip
RUN mv unzip/baysor/* .
RUN rm -rf unzip
RUN rm baysor-x86_x64-linux-v0.7.1_build.zip
ENV PATH="/app/bin:$PATH"
There was a problem hiding this comment.
The docker image from [Baysor Dockerhub](docker pull vpetukhov/baysor:v0.7.0) can also be referred to in this case
There was a problem hiding this comment.
I just checked and my image with the latest version has a compressed size of 5.41 GB, this one (vpetukhov/baysor:v0.7.0) is 2.12 GB. If this difference is not significant I would rather keep the latest version since it fixes some issues from the tool.
modules/nf-core/baysor/main.nf
Outdated
| def VERSION = "0.7.1" | ||
|
|
||
| """ | ||
| /app/bin/baysor run ${transcripts_csv} -c ${config_toml} $args |
There was a problem hiding this comment.
In the docker image, add the baysor executable to PATH
modules/nf-core/baysor/main.nf
Outdated
| when: | ||
| task.ext.when == null || task.ext.when | ||
|
|
||
| script: |
There was a problem hiding this comment.
add the following code snippet to have a check for conda
if (workflow.profile.tokenize(',').intersect(['conda', 'mamba']).size() >= 1) {
error "Baysor module does not support Conda. Please use Docker / Singularity / Podman instead."
}
Attempting to do my first module’s contribution, so some things might be improved. I also know that @khersameesh24 wanted to work on this module.
Quick note on tests:
main.nf.test.snapfile since the module’s output is non-deterministic. I was not sure on what is the best way to tackle this, so any feedback is really appreciated.