Xinglab / rmats-turbo

Other
228 stars 55 forks source link

Can not install rMATS 4.2.0 when the environment is python 3.11 #361

Open yitengfei120011 opened 9 months ago

yitengfei120011 commented 9 months ago

Can not install rMATS 4.2.0 when the environment is python 3.11 using conda install bioconda::rmats

yitengfei120011 commented 9 months ago

bioconda::rmats -> python[version='2.7.*|>=2.7,<2.8.0a0|>=3.10,<3.11.0a0|>=3.9,<3.10.0a0|>=3.8,<3.9.0a0|>=3.6,<3.7.0a0|>=3.7,<3.8.0a0']

EricKutschera commented 9 months ago

bioconda doesn't support python 3.11 yet: https://github.com/bioconda/bioconda-recipes/issues/37805

yitengfei120011 commented 9 months ago

My environment is python 3.11, which rMATS's version can support ?

EricKutschera commented 9 months ago

rMATS should work with python3.11 if you build it yourself: https://github.com/Xinglab/rmats-turbo/tree/v4.2.0#dependencies https://github.com/Xinglab/rmats-turbo/tree/v4.2.0#build

Since bioconda doesn't support python 3.11 yet, you can't use conda to install any version of rMATS to a python 3.11 environment

You could create a new environment for python 3.10 and install rMATS:

conda deactivate
conda create --prefix ./py310_rmats_env
conda activate ./py310_rmats_env
conda install -c conda-forge -c bioconda python=3.10 rmats=4.2.0
igonzals commented 9 months ago

rMATS should work with python3.11 if you build it yourself: https://github.com/Xinglab/rmats-turbo/tree/v4.2.0#dependencies https://github.com/Xinglab/rmats-turbo/tree/v4.2.0#build

Since bioconda doesn't support python 3.11 yet, you can't use conda to install any version of rMATS to a python 3.11 environment

You could create a new environment for python 3.10 and install rMATS:

conda deactivate
conda create --prefix ./py310_rmats_env
conda activate ./py310_rmats_env
conda install -c conda-forge -c bioconda python=3.10 rmats=4.2.0

Many thanks! I had the same problem and now the installation process worked!

igonzals commented 9 months ago

The installation went smoothly. However, when I try to use rmats.py I get this error:

conda activate ./py310_rmats_env python rmats-turbo/rmats.py --b1 ../../PRJNA339740/code/b1.txt -gtf ../../genomes/gencode.v45.annotation.gtf --od ../../PRJNA339740/rmats/ -t readType --nthread 12

Traceback (most recent call last): File "/kraken2/songlab/igonzalez/tools/rmats/rmats-turbo/rmats.py", line 19, in from rmatspipeline import run_pipe ModuleNotFoundError: No module named 'rmatspipeline'

Do you know how to fix it?

igonzals commented 9 months ago

@EricKutschera I just noticed that there is another rMATS folder inside py310_rmats_env/rMATS. I used that rmats.py because in that folder there was the file rmatspipeline.cpython-310-x86_64-linux-gnu.so

Now, I can work well.

yitengfei120011 commented 8 months ago

hi, I had an error installing myself and my environment is py311 ,Can you work it out for me ? image

EricKutschera commented 8 months ago

Looks like the error might be f77: Command not found

You can install gfortran and run export FC="/path/to/gfortran": https://github.com/Xinglab/rmats-turbo/blob/v4.2.0/setup_environment.sh#L27

See this post: https://github.com/Xinglab/rmats-turbo/issues/103#issuecomment-808488147

yitengfei120011 commented 8 months ago

hi, I'm quite sure of mine. bashrc document has been deleted "/home/miaoxin/Softwares/miniconda3/envs/R_4.1/lib/R/" ,why does this happen again ? Can you give you some advice to solve this problem ? image

yitengfei120011 commented 8 months ago

/home/miaoxin/Softwares/miniconda3/envs/R_4.1/lib/R/library is not mine ;

EricKutschera commented 8 months ago

That error is from the last step of build_rmats where it installs the PAIRADISE R package that rMATS uses if --paired-stats is given: https://github.com/Xinglab/rmats-turbo/blob/v4.2.0/build_rmats#L96

You can give --no-paired-model to build_rmats to skip that part of the build if you're not going to use --paired-stats

If the build is finding an R library path other than the one you want then maybe your enivornment varaibles like PATH are not set the way you expect. It sounds like you made some changes to your .bashrc. Opening a new terminal should reload your .bashrc

yitengfei120011 commented 8 months ago

image image Did not report error,but rmats.py not work . image

yitengfei120011 commented 8 months ago

image Did not report error, but not work . image

EricKutschera commented 8 months ago

No module named 'rmatspipeline' happens if python doesn't find an rmatspipeline.so file it can use. The build should have created a file like rmatspipeline.cpython-36m-x86_64-linux-gnu.so that indicates the python version used for the build. If the python version used to run rmats is different from the version used to build rmats then that "no module" error can happen. You could try using a full path to the correct version of python. If the build was run in a conda environment then make sure that environment is activated See this post: https://github.com/Xinglab/rmats-turbo/issues/251#issuecomment-1378772775

aliciaaevans commented 4 months ago

rmats on Bioconda has been built for Python 3.11 and 3.12 now.

desmodus1984 commented 4 months ago

Hi @EricKutschera

Sorry for my stupid question. I want to install rmats-Turbo to check alternative splicing events for low-high elevation populations, and I am confused what to do and how to install it. It seems that one has to build it but there is a released rmats_turbo_v4_3_0.tar.gz file, that in the user guide (https://rmats.sourceforge.io/user_guide.htm) mentions to decompress to run rmats-turbo.

I normally use conda, but I am not sure if I should make a new conda environment for rmats-turbo or if I don't/ it is unnecessary. For instance, I found that it is mentioned in conda: https://anaconda.org/bioconda/rmats; thus I would appreciate if you can confirm I can install it as mentioned - conda install bioconda::rmats - and be able to do the analysis, or if I should use ./build_rmats.

Thank you very much!

EricKutschera commented 4 months ago

You could build from the source code using ./build_rmats, but it may take some time work through any issues. Unless you have a specific reason for building from source I would recommend installing with conda. Here are commands to install and run the latest version in a new conda environment:

conda create --prefix ./rmats_conda_env
conda activate ./rmats_conda_env
conda install -c conda-forge -c bioconda rmats=4.3.0
rmats.py --help

The user guide you linked to is for an old version (v4.0.2). The latest README is on GitHub: https://github.com/Xinglab/rmats-turbo