Omicverse

Omicverse provides a Python framework for multi-omics analysis of bulk, single-cell, and spatial RNA-seq data, including tools for data deconvolution and trajectory blending to restore continuity of omitted cells in single-cell datasets.

Key Features:

• BulkTrajBlend algorithm: Leverages a Beta-Variational AutoEncoder for data deconvolution and graph neural networks for discovery of overlapping communities to interpolate and restore continuity of omitted cells within single-cell RNA sequencing datasets.
• Multi-omics analysis toolkit: Provides extensive tools for bulk RNA-seq, single-cell RNA-seq, and spatial RNA-seq analysis with seamless access to diverse methodologies.
• Data visualization: Supports data visualization for biological insights extraction.

Scientific Applications:

• Single-cell trajectory restoration: Interpolates and restores omitted cells in single-cell RNA sequencing datasets by blending information from bulk RNA-seq data.
• Bulk and single-cell RNA-seq integration: Bridges insights across bulk and single-cell sequencing to address limitations in single-cell sequencing throughput.
• Multi-omics data analysis: Enables comprehensive analysis of bulk, single-cell, and spatial RNA-seq data for biological insight extraction.

Methodology:

The BulkTrajBlend algorithm uses a Beta-Variational AutoEncoder for data deconvolution and graph neural networks for discovery of overlapping communities.