epigen / spilterlize_integrate Goto Github PK

Visualize the estimated/modeled/removed effects of batch, unwanted, and desired sources of variation.

Does not catch the error of getting input dataframe with duplicate gene/feature names till filter_features rule.

Use VOOMs inherent plotting capabilities for a diagnostic plot normVOOM_mean-variance-trend.png
https://rdrr.io/bioc/limma/man/voom.html

There is a (known) issue with installing reComBat by pip, because it lists sklearn as one of its dependencies, which is unavailable since Dec 2023.
An exisiting issue here: BorgwardtLab/reComBat#4
The underlying problem: https://github.com/scikit-learn/sklearn-pypi-package

This prevents conda from creating the reComBat environment and therefore the whole pipeline doesn't run.

this should enable to better see/spot the overall trend to judge eg filtering in case of many features (e.g., ATAC-seq data)

perform a test run with minimal configurations/options of not using a feature with e.g., empty lists []

to increase findability and application.
e.g., ChIP-seq, Methyl-Seq, miRNA-Seq, WGBS, microarray Cut&Tag ,

either from tutorials e.g., limma or wait until other modules are ready for one test that goes from SRA download until enrichment analysis -> this will produce example data for most modules.