Save raw UMI counts and metadata for each object

100%

Save raw UMI counts and metadata for each object

✓ Published0🌍 Public

Last edited May 1, 2018

Created on May 1, 2018

This example saves raw UMI counts and metadata for each timepoint in a single-cell heart maturation dataset. It iterates over six developmental stages, reading a Seurat object for each and extracting the raw UMI matrix and metadata. The code uses base R functions `as.matrix()` and `readRDS()`, then writes tab-separated files with `write.table()`. The visualization shows how raw count matrices are paired with their corresponding cell annotations, facilitating downstream differential expression analysis. The approach relies on Seurat’s stored raw data and metadata slots, producing separate DGE and metadata files per timepoint.

AI-generated description

Similar vizzes

Loading thumbnail…

Create an Rds file for the full heart maturation dataset

This example demonstrates how to convert a Seurat single-cell RNA sequencing object into a compact RDS file for a Shiny visualization app. It shows the creation of a custom S4 class that bundles tSNE coordinates, normalized expression data, cell metadata, and the top 50 differentially expressed markers per cluster. Using the dplyr and tidyr libraries for data manipulation, the script extracts the tSNE embeddings and cluster identities from the Seurat object, then merges them with metadata. The resulting clustering_module object is saved via saveRDS, enabling efficient downstream interactive exploration of heart maturation data.

FFloWuenne

72% match

Loading thumbnail…

Extract data for clustering module from all time points (For .Rdata)

This example demonstrates a data-extraction workflow that prepares Seurat clustering results from multiple developmental time points for a Shiny server application. It loads seven Seurat objects (E14.5 through P6) and, for each, extracts normalized expression matrices, t-SNE embeddings, cluster identities, and metadata. The script explicitly uses the `dplyr` and `tidyr` packages for data manipulation, converting sparse expression matrices to dense data frames and merging cell embeddings with identity assignments. The resulting data frames are saved as `.RData` files per time point, preserving the structure needed for interactive clustering visualization. The code relies on Seurat’s slot accessors (`@data`, `@dr$tsne`, `@ident`, `@meta.data`) to gather all relevant components.

FFloWuenne

67% match

Loading thumbnail…

Create shiny cluster modules for renamed datasets

This example demonstrates how to process and store single-cell RNA sequencing data across multiple developmental time points, creating Shiny-ready cluster modules for each stage. The code defines an S4 class, `clustering_module`, that bundles t-SNE embeddings, normalized expression data, metadata, and a marker list for each time point from E14.5 to P7. It loads Seurat objects, extracts the t-SNE coordinates and cell identities, filters markers by expression difference using `dplyr`’s `top_n`, and saves each time point as a compressed RDS file. The workflow uses `Seurat` for data handling and `tidyr` for reshaping, preparing interactive visualizations for the Shiny server.

FFloWuenne

67% match

Loading thumbnail…

Create .Rdata file from Seurat object for clustering module of shiny server (for .Rdata)

This example demonstrates how to convert a Seurat object into an .Rdata file for use in a clustering module of a Shiny server. The script extracts normalized expression data as a dense matrix, along with t-SNE embeddings and cell cluster identities, and merges them into a single data frame. It also includes the full metadata associated with each cell. The code uses the Seurat and dplyr APIs, specifically accessing the `@data`, `@dr$tsne@cell.embeddings`, and `@ident` slots, and saves the resulting objects with the base R `save` function.

FFloWuenne

66% match

Loading thumbnail…

Extract shiny clustering data for full dataset

This example prepares a clustered single-cell dataset for interactive visualization by extracting tSNE coordinates, cluster identities, normalized expression, and marker genes from a Seurat object. It shows how the full dataset’s clustering structure is assembled into a custom S4 object for downstream use. The code uses R with `dplyr` and `tidyr` to filter the top 50 markers per cluster by `avg_logFC`, and merges tSNE embeddings with cell identities. It saves the processed data as an RDS file for a Shiny app, relying on Seurat’s data structures and base R matrix operations.

FFloWuenne

63% match

Loading thumbnail…

Make a table of the number of cells for each group in the seurat object

The example builds a data frame of cell counts for each cluster in a Seurat object, broken down by treatment group. It first shows raw cell numbers per group per cluster, then adds normalized proportions relative to each group’s total cell count, making group compositions comparable despite unequal sample sizes. The code uses `Seurat` functions like `SubsetData` and the `@ident` and `@meta.data` slots, alongside `dplyr`’s `select` and `nrow` to count group memberships. It iterates over clusters, appending rows via `rbind`, and finally writes the result to a tab-separated file with `write.table`. The output table directly reveals per-cluster group distributions and their normalized percentages, aiding differential abundance analysis.

FFloWuenne

63% match

Loading thumbnail…

Create a cellview Rds object from a seurat expression object (updated for Seurat version 2) [MOUSE]

This example demonstrates how to prepare a Seurat single-cell expression object for use with the CellView visualization tool, specifically for mouse data. It shows the workflow of rerunning t-SNE with three dimensions, converting gene symbols to Ensembl IDs using `bitr` from clusterProfiler, and constructing the necessary `log2cpm`, `featuredata`, and `tsne.data` matrices. The code uses Seurat’s `RunTSNE`, the `Matrix` and `dplyr` packages for data manipulation, and ultimately saves the combined object as an Rds file for downstream visualization.

FFloWuenne

62% match

Loading thumbnail…

Visualize the output from GatherMolecularBarcodeDistributionByGene to assess UMI duplication rates

This example visualizes the relationship between duplicated and unique UMI counts, split by cell and by gene, to assess UMI duplication rates. Scatter plots show each cell or gene as a point, with axes for duplicated versus unique UMIs, while histograms display the distribution of mean duplication events per cell. The code uses R’s ggplot2 and plotly to build interactive scatter plots with tooltips, and dplyr for data wrangling. It calculates per-cell and per-gene duplication percentages from a table of UMI observations, highlighting overall duplication patterns.

FFloWuenne

62% match