[QUESTION] Which options are deprecated in SemiBin2?

[rdenise]

To whom it may concern,

I am reading the documentation but I am more than lost. I do not understand which file SemiBin2 requires.

From what I understand some option were mandatory for SemiBin1 but are not anymore for SemiBin2. One example is the --environment or --semi-supervised or --self-supervised. From what I understand to gain some time you should use the --environment to use the pretrained model, but what happens if you don't specify --environment. Because the documentation said that if you don't use --environment you should specify either --semi-supervised or --self-supervised, but the documentation said Use --semi-supervised to use semi-supervised learning (although that is also deprecated) so this option is deprecated, so is not used?

Also all the examples do not specify one of the 3 options so what is the default behavior of SemiBin2.

I understand that the transition between SemiBin1 to SemiBin2 introduce a lot of changes and that having a documentation the most up-to-date is time consuming, but SemiBin2 seems to be a fantastic tool and I just want to be able to use it for my data by understanding exactly which option I should use depending of my case.

For example in the option there is an option --reference-db-data-dir that said that it will download the database if not specified, but i can't afford 100Gb to be download in a random place in my computer. But the option seems deprecated, so I suppose SemiBin2 will not download it. But how can I be sure, I couldn't find in the documentation which step SemiBin2 exactly take depending on the options provided

Thank you in advance for the help,

[rdenise]

In addition the github readme and documentation are saying different thing concerning the installation to use the GPU

Documentation:

conda create -n SemiBin
conda activate SemiBin
conda install -c conda-forge -c bioconda semibin
conda install pytorch torchvision torchaudio cudatoolkit=10.2 -c pytorch-lts

Readme:

conda create -n SemiBin
conda activate SemiBin
conda install -c conda-forge -c bioconda semibin
conda install -c pytorch -c nvidia pytorch pytorch-cuda=11.8