pyRforest

pyRforest integrates Python scikit-learn RandomForestClassifier algorithms into the R environment for genomic data analysis, optimizing for computational speed and classification tasks on large genomic datasets such as RNA-seq.

Key Features:

• Random Forest Classification: Implements scikit-learn RandomForestClassifier algorithms within R for classification tasks on large genomic datasets.
• Rank-based Permutation Method: Provides a novel rank-based permutation method for biomarker identification that estimates and visualizes P-values for individual features.
• SHAP Value Calculation: Includes methods for the calculation and visualization of SHapley Additive exPlanations (SHAP) values.
• Downstream Enrichment Analysis: Supports comprehensive downstream analysis for gene ontology and pathway enrichment.

Scientific Applications:

• Genomic Data Classification: Classifies large genomic datasets, particularly those from RNA-seq experiments.
• Biomarker Identification: Identifies features with robust statistical evidence of an effect through P-value estimation and visualization.
• Model Interpretability: Improves interpretability of Random Forest models for genomic data analysis through SHAP value visualization and statistical feature ranking.

Methodology:

pyRforest integrates Python scikit-learn RandomForestClassifier algorithms into R, inheriting efficient memory management and parallelization from Python. The package employs a rank-based permutation method for biomarker identification and calculates SHapley Additive exPlanations values for feature interpretation.