DALI

How to cite

Please cite as: Verstaen K, Lammens I, Roels J, Saeys Y, Lambrecht BN, Vandamme N, Vanhee S. DALI (Diversity AnaLysis Interface): a novel tool for the integrated analysis of multimodal single cell RNAseq data and immune receptor profiling. bioRxiv 2021.12.07.471549; doi: https://doi.org/10.1101/2021.12.07.471549

Installation

To install the latest version of DALI, open R an install using devtools:

if (!requireNamespace("devtools", quietly = TRUE))
    install.packages("devtools")

devtools::install_github("vibscc/DALI")

To perform Trajectory analyses, installation of the Dynverse R packages and an installation of Docker is required. Dynverse is installed using devtools:

devtools::install_github("dynverse/dyno")

Usage

Starting the interactive shiny app

library("DALI")

## Without a seurat object
## The application will allow you to upload all necessary data via the web browser

Interactive_DALI()

## With a seurat object (1)

seuratObj <- readRDS("<path/to/seurat_object.Rds>")
Interactive_DALI(seuratObj)

(1) If this seurat object does not have the VDJ data loaded yet using Read10X_vdj(), Read10X_AIRR() or Read_AIRR(), the application will prompt you to load in the data. Select the 10X cellranger output for either the BCR and/or TCR data linked to the same gene-expression data present in your seurat object. The app will then load in the data and start up.

Loading 10X VDJ data in an existing seurat object

library("DALI")

seuratObj <- readRDS("<path/to/seurat_object.Rds>")
seuratObj <- Read10X_vdj(seuratObj, "<path/to/cellranger/bcr_or_tcr_out>", assay = "<assay>")

# <assay> can either be BCR or TCR

Example data

Example data to use with this tool can be downloaded here:
https://cloud.irc.ugent.be/public/index.php/s/9ys5czsaNtNQtSd

FAQ

Q: What cellranger folder do I need to provide to load the vdj data?

A: If you used cellranger vdj, use the folder outs from the output.
For cellranger multi, use the folder vdj_b (BCR) or vdj_t (TCR). This folder can be located in outs/per_sample_outs/<sample>.

Q: How can I use DALI on merged samples?

A: Merging of the VDJ data (TCR and/or BCR) from multiple samples is currently not supported in DALI, but will come shortly!
For now, this will require you to follow any of the following workflows:

If you used cellranger multi for the counts of the individual samples, you could aggregate the data into 1 datset using cellranger aggr. This will result in 1 count matrix and 1 set of VDJ related files which you can load into Seurat and DALI.
Manually concatenate the filtered_contig_annotation.csv or all_contig_annotation.csv. Make sure the clonotypes for each sample have a unique name! This can be done by simply appending _<SAMPLE> (replace <SAMPLE> with the actual name of your sample) to each clonotype definition (column raw_clonotype_id). This is necessary because clonotype 1 of sample 1 has no relation to clonotype 1 of sample 2. You could also define the clonotypes again on the combined dataset to find some common clonotypes between different samples. This could be done with tools like immcantation

Q: How can I process gamma/delta TCR data (gdTCR) using cellranger

A: Cellranger doesn't officially support this type of data (yet). There are however some instructions here on how to work around the cellranger limitations.
NOTE: Be careful if you want to load data into DALI processed thsi way if you already have 'normal' TCR (alpha + beta) data in your object. Make sure to change the naming of the V,D,J and C region back to its original name in the output files used by DALI: airr_rearrangement.tsv and filtered/all_contig_annotations.csv