Usage of functions

In this tutorial we compile examples on how to use the different functions available in the R package. You can refer to previous tutorial to see how to use CellBender and perform quality control and basic downstream analysis of sc/snRNA-seq. We are going to use the object generated in the section: Quality control of sc/snRNA-seq.

Libraries & Set-Up

#> 2025-06-26 17:19:30 - Using existing conda environment at: /home/mariano/.venv/DOtools
#> 2025-06-26 17:19:30 - Python packages ready for DOtools!

Dotplot

The DO.Dotplot function generates a plot in a similar manner as the Dotplot function implemented in scanpy but with additional functionalities, e.g. the visualisation of 3 variables at the same time. For example, we can visualise the expression of a gene across celltypes and conditions:

DO.Dotplot(Seu_object = Seu_obj,
           group.by.x = "condition",
           group.by.y = "annotation",
           Feature = "NKG7")
#> Scale for size is already present.
#> Adding another scale for size, which will replace the existing scale.

UMAP

The DO.UMAP function facilitatesthe visualisation of UMAP embeddings since it combines the ability to show cell groupings and gene expression.


DO.UMAP(Seu_object = Seu_obj,
        group.by = "annotation",
        split.by = "condition",
        FeaturePlot = T,
        features = "NKG7",
        legend.position = "right")

Cell composition changes

As shown in the tutorial, we can also test for significant changes in cell population.


DO.CellComposition(Seu_object = Seu_obj,
                   assay_normalized = "RNA",
                   cluster_column = "annotation",
                   sample_column = "orig.ident",
                   condition_column = "condition",
                   transform_method = "arcsin",
                   n_reps = 3)
#> ℹ Using the 'counts' assay as the X matrix
#> [INFO] Your data doesn't have replicates! Artificial replicates will be simulated to run scanpro.
#> [INFO] Simulation may take some minutes...
#> [INFO] Generating 3 replicates and running 100 simulations...
#> [INFO] Finished 100 simulations in 2.26 seconds
#> Using orig.ident, condition as id variables
#> Using condition as id variables

Expression of genes and continuous metadata

We can visualize the average expression of a gene in a cell type or continuous metadata information across condition with violinplots, barplots and boxplots. Additionally, we can test for significance.

p <- DO.VlnPlot(Seu_obj,
                Feature = "NKG7",
                group.by = "condition",
                group.by.2 = "annotation",
                ctrl.condition = "healthy")
#> Using group, orig.ident, annotation as id variables
#> ListTest empty, comparing every sample with each other

Seu_obj_NK <- DO.Subset(Seu_obj, ident = "annotation", ident_name = "NK")
#> Specified 'ident_name': expecting a categorical variable.
DO.BarplotWilcox(Seu_obj_NK,
                 group.by = "condition",
                 ctrl.condition = "healthy",
                 Feature = "NKG7",
                 x_label_rotation = 0)
#> Using condition, orig.ident as id variables
#> ListTest empty, comparing every sample with each other

set.seed(123)
Seu_obj$rdm_sample <- sample(rep(c("A", "B", "C"), length.out = ncol(Seu_obj)))
DO.BoxPlot(Seu_obj,
           group.by = "rdm_sample",
           ctrl.condition = "A",
           Feature = "nCount_RNA",
           step_mod = 100,
           plot_sample = F)
#> Using group, cluster as id variables
#> ListTest empty, comparing every sample with each other
#> Scale for fill is already present.
#> Adding another scale for fill, which will replace the existing scale.

Session information