PathwayForte

Python package for pathway database benchmarking.

A Python package for benchmarking pathway databases with functional enrichment and prediction methods tasks.

Command Line Interface :

Mappings

Installation Current version on PyPI Stable Supported Python Versions Apache-2.0

pathway_forte can be installed from PyPI with the following command in your terminal:

$ python3 -m pip install pathway_forte

The latest code can be installed from GitHub with:

$ python3 -m pip install git+https://github.com/pathwayforte/pathway-forte.git

For developers, the code can be installed with:

$ git clone https://github.com/pathwayforte/pathway-forte.git
$ cd pathway-forte
$ python3 -m pip install -e .

Main Commands

The table below lists the main commands of PathwayForte.

Command

Action

datasets

Lists of Cancer Datasets

export

Export Gene Sets using ComPath

ora

List of ORA Analyses

fcs

List of FCS Analyses

prediction

List of Prediction Methods

Functional Enrichment Methods

  • ora. Lists Over-Representation Analyses (e.g., one-tailed hyper-geometric test).

  • fcs. Lists Functional Class Score Analyses such as GSEA and ssGSEA using GSEAPy.

Prediction Methods

pathway_forte enables three classification methods (i.e., binary classification, training SVMs for multi-classification tasks, or survival analysis) using individualized pathway activity scores. The scores can be calculated from any pathway with a variety of tools (see 1) using any pathway database that enables to export its gene sets.

  • binary. Trains an elastic net model for a binary classification task (e.g., tumor vs. normal patients). The training is conducted using a nested cross validation approach (the number of cross validation in both loops can be selected). The model used can be easily changed since most of the models in scikit-learn (the machine learning library used by this package) required the same input.

  • subtype. Trains a SVM model for a multi-class classification task (e.g., predict tumor subtypes). The training is conducted using a nested cross validation approach (the number of cross validation in both loops can be selected). Similarly as the previous classification task, other models can quickly be implemented.

  • survival. Trains a Cox’s proportional hazard’s model with elastic net penalty. The training is conducted using a nested cross validation approach with a grid search in the inner loop. This analysis requires pathway activity scores, patient classes and lifetime patient information.

Other

  • export. Export GMT files with current gene sets for the pathway databases included in ComPath 2.

  • datasets. Lists the TCGA data sets 3 that are ready to run in pathway_forte.

References

1

Lim, S., et al. (2018). Comprehensive and critical evaluation of individualized pathway activity measurement tools on pan-cancer data. Briefings in bioinformatics, bby125.

2

Domingo-Fernández, D., et al. (2018). ComPath: An ecosystem for exploring, analyzing, and curating mappings across pathway databases. npj Syst Biol Appl., 4(1):43.

3

Weinstein, J. N., et al. (2013). The cancer genome atlas pan-cancer analysis project. Nature genetics, 45(10), 1113.

Indices and Tables