MineICA

MineICA performs integrative Independent Component Analysis (ICA) on multiple transcriptome datasets to extract statistically independent gene expression components and associate them with molecular, clinical, and pathological annotations.

Key Features:

• Independent Component Analysis (ICA): Extracts statistically independent components from large-scale transcriptome data to identify underlying biological signals.
• Integrative Data Approach: Integrates molecular, clinical, and pathological data with transcriptomic profiles to link components to sample annotations and gene sets.
• Association Studies: Computes associations between ICA components and sample annotations or gene sets to enhance biological interpretability.
• Comparative Analysis: Compares independent components across datasets using correlation-based graph techniques to assess similarities and differences between studies.

Scientific Applications:

• Feature Extraction: Transforms large-scale gene expression data into reduced component representations for downstream analysis and interpretation.
• Data Reduction and Analysis: Reduces dataset dimensionality to facilitate identification of biomarkers and investigation of disease mechanisms.

Methodology:

MineICA applies Independent Component Analysis (ICA) to decompose transcriptome data into statistically independent sources, integrates molecular, clinical, and pathological annotations with the resulting components, and uses correlation-based graph methods to compare components across datasets.