Can you share the already built images?

[JiahuaQu]

Hi, we are learning from this project. Because our HPC does not give us the root authority to use Docker, we have to either build a Singularity image based on a Docker image, or build the Docker image in our local computer. We are trying doing it in our local computer, but it has been taking too much time and has not been completed. Finally, it failed to build the ripseq image. As you must have the already built Docker images, can you share them on Docker hub or in any other way, like share the Singularity images directly? So that we can build our own Singularity image on our HPC. Thank you.

[alnfedorov]

Hi! Unfortunately, original images didn't survive through all this time, and it was clearly a mistake not to publish it originally. As you noted, the rip-seq image is pretty much useless now since building it takes eternity (left it to run for about a week with no success). Could you try using the following Dockerfile?

FROM mambaorg/micromamba:1.5.6

RUN micromamba config append channels defaults && \
    micromamba config append channels bioconda && \
    micromamba config append channels conda-forge && \
    micromamba install -y -n base \
        r-base=4.1.1 \
        bioconductor-deseq2=1.32.0 \
        bioconductor-apeglm=1.14.0 \
        bioconductor-tximport=1.20.0 \
        bioconductor-degreport=1.28.0 \
        pybedtools=0.8.2 \
        deeptools=3.5.1 \
        seaborn=0.11.2 \
        pandas=1.2.4 \
        matplotlib=3.4.1 \
        joblib=0.14.1 \
        biopython=1.76 \
        python=3.8 && \
    micromamba clean --all --yes

ARG MAMBA_DOCKERFILE_ACTIVATE=1
RUN python3.8 -m pip install htseq==0.12.3

WORKDIR /project

This hasn't been tested, but it should be nearly identical to the original one. Alternatively, here are the links to Docker Hub: chip-seq and rip-seq (built from the file above).

[JiahuaQu]

Thank you so much. I will give them a try.

I was trying the rip-seq image, but the program showed that the python was not found inside the image?

Singularity> python bash: python: command not found

I listed and exported all installed packages in the attached file. Can you have a check of the file and the image? Thank you. packages.txt

And I continued to try the chipseq image, but the snakemake can't be found. bash: snakemake: command not found

[alnfedorov]

I was trying the rip-seq image, but the program showed that the python was not found inside the image?

Singularity> python bash: python: command not found

That's strange. Maybe try python3.8 or python3? It should be there since it's used in the Dockerfile.

And I continued to try the chipseq image, but the snakemake can't be found. bash: snakemake: command not found

Note, there is a separate Dockerfile for the data preprocessing (called the pipeline). It's available here.

I thought the problem was only with the analysis step, so I built and uploaded images only for the analysis step. I will take a look to see if the chip pipeline image is buildable.

[alnfedorov]

When running locally, python is available in the rip-seq environment: Not sure why it's not available via Singularity...

[JiahuaQu]

Thank you for the check. Then I will try to use Docker in my local. With regard to the chip-seq preprocess pipeline, I used to build it, but it always showed some module broken, particularly those in multiqc. Therefore, hope you can help with that pipeline as well.

[alnfedorov]

Here is a pre-built ChIP-seq pipeline image. I might be able to help with troubleshooting if you provide more details, but I would strongly recommend using a modern and well-supported Nextflow alternative (e.g., nfcore).

[JiahuaQu]

Thank you. Previously, when I built and used the Chipseq image, the pipeline was always terminated with the error:

[ERROR ] multiqc : Oops! The 'custom_content' MultiQC module broke... [ERROR ] multiqc : Oops! The 'ngsderive' MultiQC module broke... [ERROR ] multiqc : Oops! The 'ngsderive' MultiQC module broke... [ERROR ] multiqc : Oops! The 'purple' MultiQC module broke... [ERROR ] multiqc : Oops! The 'conpair' MultiQC module broke... [ERROR ] multiqc : Oops! The 'peddy' MultiQC module broke... [ERROR ] multiqc : Oops! The 'conpair' MultiQC module broke... ... [ERROR ] multiqc : Oops! The 'quast' MultiQC module broke...

OR: File "/opt/conda/lib/python3.7/site-packages/markdown/util.py", line 32, in from typing import TYPE_CHECKING, Generic, Iterator, NamedTuple, TypeVar, TypedDict, overload ImportError: cannot import name 'TypedDict' from 'typing' (/opt/conda/lib/python3.7/typing.py)