GReNaDIne

GReNaDIne infers gene regulatory networks (GRNs) from high-throughput gene expression data to enable analysis of gene regulatory mechanisms.

Key Features:

• Diverse Inference Methods: Implements 18 distinct data-driven gene regulatory network inference algorithms.
• Preprocessing Techniques: Provides eight generalist preprocessing techniques applicable to RNA-seq and microarray datasets.
• RNA-seq Normalization Methods: Includes four normalization methods specifically tailored for RNA-seq data.
• Ensemble Inference Capability: Supports combination of results from multiple inference tools to create robust ensembles.
• Benchmark Validation: Has been assessed using the DREAM5 challenge benchmark dataset.
• Implementation and Modularity: Implemented in Python with a modular design that supports integrated preprocessing and postprocessing.
• Compatibility with PYSCENIC: Produces outputs compatible with complementary refinement tools such as PYSCENIC.

Scientific Applications:

• GRN Reconstruction: Inferring gene regulatory networks from high-throughput gene expression data, including RNA-seq and microarray.
• Deciphering Gene Interactions: Deciphering complex gene interactions within cellular processes for systems biology studies.
• Method Benchmarking and Selection: Benchmarking and comparative evaluation of inference algorithms using the DREAM5 dataset.
• Downstream Refinement: Generating outputs suitable for downstream refinement and interpretation with tools like PYSCENIC.

Methodology:

Implemented in Python and providing eight preprocessing techniques, four RNA-seq normalization methods, 18 inference algorithms, ensemble combination of inference results, modular preprocessing/postprocessing, and validation on the DREAM5 benchmark.