phantasus

Phantasus provides interactive exploratory and statistical analysis of gene expression datasets to support transcriptomic profiling, dataset quality assessment, and differential expression testing.

Key Features:

• Public dataset access: Provides access to more than 84,000 public gene expression datasets for large-scale analyses.
• User-uploaded data: Supports analysis of user-uploaded expression matrices.
• Heatmap engine (Morpheus): Uses a JavaScript heatmap engine derived from Morpheus for large-scale heatmap visualization of expression matrices.
• R/Bioconductor integration: Integrates with R and Bioconductor via the OpenCPU API to execute server-side analytical routines.
• Analytical methods: Implements normalization, filtering, k-means clustering, principal component analysis, and differential expression analysis using the limma package.
• End-to-end workflow support: Supports dataset loading, quality assessment, differential expression testing, and downstream interpretation.
• Hybrid architecture: Combines client-side interactive visualization with server-side R/Bioconductor computation via OpenCPU.

Scientific Applications:

• Exploratory transcriptomics: Interactive inspection and transformation of gene expression matrices for hypothesis generation.
• Quality assessment: Assessment of expression-matrix quality prior to downstream analyses.
• Differential expression testing: Identification of differentially expressed genes using the limma package.
• Clustering and dimensionality reduction: Pattern discovery in gene expression via k-means clustering and principal component analysis.
• Large-scale/public compendium analysis: Comparative and integrative analyses across >84,000 public datasets and user-uploaded data.
• Downstream interpretation: Preparation of results for biological interpretation following differential expression testing.

Methodology:

Uses a JavaScript heatmap engine derived from Morpheus; connects to R/Bioconductor via the OpenCPU API; performs normalization, filtering, k-means clustering, principal component analysis, and differential expression analysis using the limma package; supports dataset loading and quality assessment.