Linnorm

Linnorm normalizes and transforms high-throughput count data (RNA-seq, single-cell RNA-seq, ChIP-seq) in R to enable parametric statistical tests and downstream analyses within the Bioconductor ecosystem.

Key Features:

• Data transformation for parametric tests: Transforms raw count data to a distribution more suitable for parametric statistical testing.
• Cell subpopulation analysis and visualization: Uses PCA-based clustering to identify and visualize distinct cell subpopulations in complex datasets.
• Differential expression and peak detection: Performs differential expression analysis using the limma package for RNA-seq and supports differential peak detection for ChIP-seq.
• Highly variable gene discovery and visualization: Identifies highly variable genes across samples and provides visualization procedures to explore them.
• Gene correlation network analysis and visualization: Constructs and visualizes gene correlation networks to examine co-expression patterns.
• Hierarchical clustering and plotting: Applies hierarchical clustering methods and generates plots to display sample or feature groupings.

Scientific Applications:

• Genomics research: Supports gene expression profiling and chromatin accessibility analyses for genomic studies.
• Molecular biology investigations: Facilitates analysis of molecular mechanisms and differential expression relevant to biological processes and diseases.
• Single-cell analysis: Enables normalization, clustering, and variability analyses for single-cell RNA-seq to investigate cellular heterogeneity.

Methodology:

Implemented as an R package within the Bioconductor project and interoperable with other Bioconductor packages; methods explicitly include count-data transformation, PCA-based clustering, limma-based differential expression, differential peak detection for ChIP-seq, highly variable gene identification, gene correlation network construction, and hierarchical clustering.