Problems installing PlasClass

[alanxelena]

Hi!

I'm attempting to install PlasClass in a VM running ubuntu 20.04. The installation instructions say that install.py will automatically install the dependencies (NumPy and Scikit-learn) but it somehow fails. I can successfully install Scikit-learn independently but I get the following message related to PlasClass:

ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
plasclass-dpellow 0.1 requires joblib==1.2.0, but you have joblib 1.1.0 which is incompatible.
plasclass-dpellow 0.1 requires numpy==1.22.0, but you have numpy 1.22.4 which is incompatible.
plasclass-dpellow 0.1 requires scikit-learn==0.21.3, but you have scikit-learn 1.1.2 which is incompatible.
plasclass-dpellow 0.1 requires scipy==1.3.2, but you have scipy 1.8.1 which is incompatible.

It seems to me that asides from joblib, every version is newer than the required. Is this a bug in PlasClass?

Any help would be really appreciated.

[dpellow]

Thanks for the update. Yes, the required versions are older as the tool was developed a few years ago. Are you trying to install in a virtual environment? It should automatically install the correct versions. It may also be easier to install through bioconda - the versions will be handled automatically.

[alanxelena]

Hey! Thanks for the reply. I'm installing it on a server based Ubuntu machine, I had some problems with bioconda in the past that's why I tried to avoid going that way. Is there a way to modify PlasClass to recognise versions newer than expected as valid? I'm guessing the base functionalities of each tool should remain the same despite the updates. Anyhow, if there isn't a workaround I might give bioconda another shot.

Thanks again!

[dpellow]

you should be able to modify the setup.py file to the versions you have, but I'm not sure that the dependencies will all be correct or if any of the updated versions contain modifications that are not compatible with some parts of the code. It should all be taken care of in bioconda, but let me know otherwise. If you are installing in a clean virtual environment it also shouldn't be an issue, did you try this?