Overview

Single-Cell RNA-Seq Data Analysis and Mining Platform

BxGenomics scRNA-Seq View can be used for both private single-cell data and public datasets. Private data are analyzed by our best-practice pipeline. Public data are either provided by the authors or have been analyzed and curated using our pipeline.

The system has a strong scientific basis, with its core components published or submitted in peer-reviewed journals.

The single-cell RNA-Seq analysis pipeline integrates several popular tools (Seurat, Scanpy, Harmony, LIGER) and provides a semi-automated workflow for 1) processing of raw data, 2) batch-correction with several harmonization methods, 3) reference-based cell type label transfer, and 4) multi-sample multi-condition single-cell level differential gene expression analysis. The results can be directly loaded into the BxGenomics scRNA-Seq web databases and downstream analysis and visualization.

The workflow for single-cell RNA-Seq data analysis and visualization in BxGenomics.

The integrated BxGenomics web database includes single-cell RNA-Seq results from over 400 published datasets, as well as each user's private datasets. It provides an efficient data management system for single-cell datasets. Users can search the database using multiple attributes such as publication, keywords, species, and cell types. Advanced filters and tools for gene expression across datasets enable users to quickly find the datasets of interest and compare the expression of genes across them.

Importantly, each dataset can be loaded into the cellxgene VIP tool, which facilitates the rapid exploration of scRNA-Seq results in an interactive and scalable manner. This tool empowers researchers to gain insights into cell composition, gene expression profiles, and differentially expressed genes across cell types. By leveraging over 20 commonly used plotting functions and high-level analysis techniques in single-cell research, researchers can refine their analysis and deepen their understanding of the data.

Unique to BxGenomics, scRNA-Seq results are visualized using the powerful cellxgene VIP system. Cellxgene VIP includes analytical modules that provide essential functions for interactive visualization and generation of publication-ready plots. These modules include: (a) Multi-tSNE/UMAP plots that visually highlight which cells express cell markers on selected embeddings (UMAP based on harmony batch correction in this example). (b) Dual-gene plots that highlight cells expressing SYT1 and GAD1 (green SYT1 only, red GAD1 only, yellow co-expression of STY1 and GAD1) with an expression cutoff of 2.2. (c) Stacked barplots that demonstrate the fraction of each major cell type across each sample (C represents Control and MS represents MS patients). (d) Trackplots that show the expression of lineage marker genes across individual cells in annotated clusters. (e) Violin plots that show the AQP4 gene expression across cell types. (f) Sankey diagrams (also known as Riverplots) that provide quick and easy ways to explore the interdependent relationship of variables in the MS snRNAseq datasets. (g) Density plots that show the expression of marker genes across annotated clusters and split across cell types. (h) Stacked violin and dot plots are the key visualizations of selected cell markers across cell types. They highlight the selective expressions, and validate the scRNA-Seq approaches and visualization methods. (i) Command Line Interface (CLI) exposed by mini Jupyter Notebook that provides maximal flexibility in doing various analyses overall or sliced single-cell datasets.