scPCA

scPCA integrates sparse principal component analysis (sPCA) with contrastive principal component analysis (cPCA) to extract sparse, biologically relevant signals from high-dimensional biological datasets, including those generated by high-throughput sequencing, by leveraging control data to separate technical noise from biological variation.

Key Features:

• Sparse Principal Component Analysis (sPCA): Implements sPCA to produce sparse components that enhance interpretability of high-dimensional biological data.
• Contrastive PCA (cPCA): Utilizes control data in cPCA to contrast target and control samples and separate technical noise from genuine biological signals.
• Stability and Interpretability: Emphasizes stability of extracted components to improve reproducibility and biological interpretability.
• Relevance and Signal Recovery: Recovers features that capture biologically relevant variation while suppressing unwanted technical variation.

Scientific Applications:

• Protein expression datasets: Applied to protein expression datasets to identify salient expression patterns amid high dimensionality.
• Microarray gene expression profiles: Applied to microarray gene expression profiles for dimensionality reduction and extraction of relevant signals.
• Single-cell transcriptome sequencing data: Applied to single-cell transcriptome sequencing data to extract sparse signals from noisy single-cell measurements.

Methodology:

Combines sPCA and cPCA and integrates control data to remove unwanted variation and extract sparse components.