In lahuuki/DeconvoBuddies: Helper Functions for LIBD Deconvolution

knitr::opts_chunk$set(
    collapse = TRUE,
    comment = "#>",
    crop = NULL ## Related to https://stat.ethz.ch/pipermail/bioc-devel/2020-April/016656.html
)

## Track time spent on making the vignette
startTime <- Sys.time()

## Bib setup
library("RefManageR")

## Write bibliography information
bib <- c(
    R = citation(),
    BiocStyle = citation("BiocStyle")[1],
    knitr = citation("knitr")[1],
    RefManageR = citation("RefManageR")[1],
    rmarkdown = citation("rmarkdown")[1],
    sessioninfo = citation("sessioninfo")[1],
    testthat = citation("testthat")[1],
    DeconvoBuddies = citation("DeconvoBuddies")[1],
    DeconvoBuddiespaper = citation("DeconvoBuddies")[2]
)

Basics

Install DeconvoBuddies

R is an open-source statistical environment which can be easily modified to enhance its functionality via packages. r Biocpkg("DeconvoBuddies") is a R package available via the Bioconductor repository for packages. R can be installed on any operating system from CRAN after which you can install r Biocpkg("DeconvoBuddies") by using the following commands in your R session:

if (!requireNamespace("BiocManager", quietly = TRUE)) {
    install.packages("BiocManager")
}

BiocManager::install("DeconvoBuddies")

## Check that you have a valid Bioconductor installation
BiocManager::valid()

Required knowledge

r Biocpkg("DeconvoBuddies") is based on many other packages and in particular in those that have implemented the infrastructure needed for dealing with snRNA-seq data. That is, packages like r Biocpkg("SingleCellExperiment").

If you are asking yourself the question "Where do I start using Bioconductor?" you might be interested in this blog post.

Asking for help

As package developers, we try to explain clearly how to use our packages and in which order to use the functions. But R and Bioconductor have a steep learning curve so it is critical to learn where to ask for help. The blog post quoted above mentions some but we would like to highlight the Bioconductor support site as the main resource for getting help: remember to use the DeconvoBuddies tag and check the older posts. Other alternatives are available such as creating GitHub issues and tweeting. However, please note that if you want to receive help you should adhere to the posting guidelines. It is particularly critical that you provide a small reproducible example and your session information so package developers can track down the source of the error.

Citing DeconvoBuddies

We hope that r Biocpkg("DeconvoBuddies") will be useful for your research. Please use the following information to cite the package and the overall approach. Thank you!

## Citation info
citation("DeconvoBuddies")

Quick start to using DeconvoBuddies

Let's load some packages we'll use in this vignette.

suppressMessages({
    library("DeconvoBuddies")
    library("SummarizedExperiment")
    library("dplyr")
    library("tidyr")
    library("tibble")
})

Access Data

Use fetch_deconvo_data Download RNA sequencing data from the Human DLPFC r Citep(bib[["DeconvoBuddiespaper"]]).

• rse_gene: 110 samples of bulk RNA-seq. [110 bulk RNA-seq samples x 21k genes] (41 MB).

• sce : snRNA-seq data from the Human DLPFC. [77k nuclei x 36k genes] (172 MB)

• sce_DLPFC_example: Sub-set of sce useful for testing. [10k nuclei x 557 genes] (49 MB)

## Access and snRNA-seq example data
if (!exists("sce_DLPFC_example")) sce_DLPFC_example <- fetch_deconvo_data("sce_DLPFC_example")

## Explore snRNA-seq data in sce_DLPFC_example
sce_DLPFC_example

## Access Bulk RNA-seq data
if (!exists("rse_gene")) rse_gene <- fetch_deconvo_data("rse_gene")

## Explore bulk data in rse_gene
rse_gene

For more details on this dataset, and an example deconvolution run check out the Vignette: Deconvolution Benchmark in Human DLPFC.

Marker Finding

Using MeanRatio to Find Cell Type Markers

Accurate deconvolution requires highly specific marker genes for each cell type to be defined. To select genes specific for each cell type, you can evaluate the MeanRatio for each gene x each cell type, where MeanRatio = mean(Expression of target cell type)/mean(Expression of highest non-target cell type). These values can be calculated for a single cell RNA-seq dataset using get_mean_ratio2().

## find marker genes with get_mean_ratio
marker_stats <- get_mean_ratio(sce_DLPFC_example,
                               cellType_col = "cellType_broad_hc", 
                               gene_name = "gene_name", 
                               gene_ensembl = "gene_id")

## explore tibble output, gene with high MeanRatio values are good marker genes
marker_stats

For more discussion of finding marker genes with DeconvoBuddies check out the Vignette: Finding Marker Genes with DeconvoBuddies.

Plotting Tools

Creating A Cell Type Color Pallet

As you work with single-cell data and deconovoltion outputs, it is very useful to establish a consistent color pallet to use across different plots. The function create_cell_colors() returns a named vector of hex values, corresponding to the names of cell types. This list is compatible with functions like ggplot2::scale_color_manual().

There are three pallets to choose from to generate colors:

• "classic" (default): Set1 from RColorBrewer - max 9 colors

• "gg": Equi-distant hues, same process for selecting colors as ggplot - no maximum number

• "tableau": tableau20 color set - max 20 colors

test_cell_types <- c("cell_A", "cell_B", "cell_C", "cell_D", "cell_E")

test_cell_colors_classic <- create_cell_colors(cell_types = test_cell_types, pallet = "classic", preview = TRUE)
test_cell_colors_gg <- create_cell_colors(cell_types = test_cell_types, pallet = "gg", preview = TRUE)
test_cell_colors_tableau <- create_cell_colors(cell_types = test_cell_types, pallet = "tableau", preview = TRUE)

test_cell_colors_tableau

If there are sub-cell types with consistent delimiters, the split argument creates a scale of related colors. This helps expand on the maximum number of colors and makes your pallet flexible when considering different 'resolutions' of cell types.

my_cell_types <- levels(sce_DLPFC_example$cellType_hc)
my_cell_colors <- create_cell_colors(
    cell_types = my_cell_types,
    pallet = "classic",
    preview = TRUE,
    split = "_"
)

Plot Expression of Top Markers

The function plot_marker_express() helps quickly visualize expression of top marker genes, by ordering and annotating violin plots of expression over cell type. Here we'll plot the expression of the top 6 marker genes for Astrocytes.

# plot expression of the top 6 Astro marker genes
plot_marker_express(
    sce = sce_DLPFC_example,
    stats = marker_stats,
    cell_type = "Astro",
    n_genes = 6,
    cellType_col = "cellType_broad_hc",
    color_pal = my_cell_colors
)

The violin plots of gene expression confirm the cell type specificity of these marker genes, most of the nueli with high expression of these six genes are Astro.

Plot Composition Bar Plot

The output of deconvolution are cell type estimates that sum to 1. A good visulization for these predictions is a stacked bar plot. The function plot_composition_bar() creates a stacked bar plot showing the cell type proportion for each sample, or the average proportion for a group of samples.

# access the colData of a test rse dataset
pd <- colData(rse_bulk_test) |>
    as.data.frame()

## pivot data to long format and join with test estimated proportion data
est_prop_long <- est_prop |>
    rownames_to_column("RNum") |>
    pivot_longer(!RNum, names_to = "cell_type", values_to = "prop") |>
    left_join(pd) 

## explore est_prop_long
est_prop_long

## the composition bar plot shows cell type composition for Sample
plot_composition_bar(est_prop_long, x_col = "RNum",
                     add_text = FALSE) +
    ggplot2::scale_fill_manual(values = test_cell_colors_classic)

## the composition bar plot shows the average cell type composition for each Dx
plot_composition_bar(est_prop_long, x_col = "Dx") +
    ggplot2::scale_fill_manual(values = test_cell_colors_classic)

Reproducibility

The r Biocpkg("DeconvoBuddies") package r Citep(bib[["DeconvoBuddies"]]) was made possible thanks to:

• R r Citep(bib[["R"]])
• r Biocpkg("BiocStyle") r Citep(bib[["BiocStyle"]])
• r CRANpkg("knitr") r Citep(bib[["knitr"]])
• r CRANpkg("RefManageR") r Citep(bib[["RefManageR"]])
• r CRANpkg("rmarkdown") r Citep(bib[["rmarkdown"]])
• r CRANpkg("sessioninfo") r Citep(bib[["sessioninfo"]])
• r CRANpkg("testthat") r Citep(bib[["testthat"]])

This package was developed using r BiocStyle::Biocpkg("biocthis").

Code for creating the vignette

## Create the vignette
library("rmarkdown")
system.time(render("DeconvoBuddies.Rmd", "BiocStyle::html_document"))

## Extract the R code
library("knitr")
knit("DeconvoBuddies.Rmd", tangle = TRUE)

Date the vignette was generated.

## Date the vignette was generated
Sys.time()

Wallclock time spent generating the vignette.

## Processing time in seconds
totalTime <- diff(c(startTime, Sys.time()))
round(totalTime, digits = 3)

R session information.

## Session info
library("sessioninfo")
options(width = 120)
session_info()

Bibliography

This vignette was generated using r Biocpkg("BiocStyle") r Citep(bib[["BiocStyle"]]) with r CRANpkg("knitr") r Citep(bib[["knitr"]]) and r CRANpkg("rmarkdown") r Citep(bib[["rmarkdown"]]) running behind the scenes.

Citations made with r CRANpkg("RefManageR") r Citep(bib[["RefManageR"]]).

## Print bibliography
PrintBibliography(bib, .opts = list(hyperlink = "to.doc", style = "html"))

lahuuki/DeconvoBuddies documentation built on Aug. 12, 2024, 8:38 p.m.