Analysis of an unknown single-cell dataset

Analysis of an unknown single-cell dataset for Single Cell and Spatial Omics class of the QCB master

Analysis Workflow

1. Quality Control and Filtering

• QC Metrics: Evaluate number of genes detected per cell, total counts per cell, and percentage of reads mapping to mitochondrial genes or spike-in RNAs (ERCCs).
• Filtering: Remove cells with low counts, low gene detection, or high mitochondrial content (or high ERCC counts).

2. Normalization and Scaling

• Normalization: Apply LogNormalize to scale data, ensuring each cell has the same total expression.
• Variable Feature Identification: Use FindVariableFeatures to identify highly variable genes.
• Scaling: Standardize data to zero mean and unit variance using ScaleData.

3. Dimensionality Reduction

• PCA: Perform Principal Component Analysis to reduce data dimensionality.
• UMAP/t-SNE: Further reduce dimensions for visualization using UMAP or t-SNE.

4. Clustering

• Clustering: Group cells into clusters using the Louvain or Leiden algorithm based on gene expression patterns.

5. Identification of Marker Genes

• Marker Genes: Conduct differential expression analysis (FindMarkers) to identify cluster-specific marker genes.

6. Cell Cycle Analysis

• Cell Cycle Scoring: Assign cell cycle phases (G1, S, G2/M) to each cell using known cell cycle markers.

7. Annotation

• Cluster Annotation: Annotate clusters based on known marker genes or databases (e.g., CellMarker, PanglaoDB).
• Hypothesis Formation: Formulate a hypothesis about the tissue origin and cell types based on clusters and marker genes.

Credits

This analysis was conducted using the following R packages:

• Seurat: An R package designed for QC, analysis, and exploration of single-cell RNA-seq data.
• ggplot2: An R package for data visualization based on the grammar of graphics.
• tidyverse: A collection of R packages for data manipulation and visualization, including ggplot2.
• patchwork: An R package to combine separate ggplots into a single graphic.
• HGNChelper: Functions for identifying and correcting HGNC human gene symbols and MGI mouse gene symbols.
• ggraph: Visualization of graph networks using ggplot2.
• igraph: Network analysis and visualization.
• data.tree: Create and manipulate tree structures from hierarchical data.