failing to run the initial example

[deniskristak]

Hi,
I've been trying to build this package from source.
Because this is part of work on Easybuild HPC package manager I am also taking care of dependencies manually, not using conda.

I'm using tag 0.2.0.

It seems to install fine, but currently, when I try to run the initial example snakemake -n --cores 1 --configfile config.yaml,
I am getting this error:

Building DAG of jobs...
MissingInputException in rule integration_prepare in file /kyukon/scratch/gent/vo/001/gvo00117/easybuild/RHEL8/haswell-ib/software/scib-pipeline/0.2.0-foss-2022a/pipeline/Snakefile, line 32:
Missing input files for rule integration_prepare:
    output: /storage/groups/ml01/workspace/scIB/pancreas/prepare/unscaled/hvg/adata_pre.h5ad
    wildcards: scenario=pancreas, scaling=unscaled, hvg=hvg, prep=h5ad
    affected files:
        /storage/groups/ml01/workspace/maren.buettner/data_integration/data/human_pancreas/human_pancreas_norm.h5ad

Does it ring any bells?
Thank you very much.

[LuckyMD]

Hi @deniskristak, it sounds like you need to adapt the config file you are using to point this towards the data you want to run this with.

[deniskristak]

Hi @LuckyMD,
thanks for catching up with me so early. Can you point me to where I could find the data you're talking about?
I honestly don't really know what kind of data the software is working with... I am only trying to run some sort of initial example according to the documentation, so that I know it's not throwing any errors or anything and that it installs all right.
Thanks!

[LuckyMD]

Hi @deniskristak,

Sure, the data is on Figshare as mentioned in the README. We also have a test run of the pipeline though that @mumichae added. She can tell you more about that feature so you don't have to download the data separately.

[mumichae]

Hi, for a test datasets just for implementation testing I'd recommend you check out the data directory of the repo. There should be a script that you can use to generate the test data as described here. You need to add the real path of the generated object to the config file for the pipeline to work.

[deniskristak]

this did help me progress, thanks! turns out I was using the latest official release you guys have (maybe think about adding next one? )