Inference of miRNA‐target interactions under different conditions

For scRNA-seq datasets with different conditions, we allow users to infer the cell-cell communication mediated by the miRNA-target interactions with the parameter condition in the function create_miRTalk. Also, users can perform the inference for each condition, seperately. Given the different highly varible genes with find_hvtg between the inference of different conditions together and seperately, the results might be slightly different. Step-by-step procedures are shown below:

[1] load the example data

> load(paste0(system.file(package = 'miRTalk'), "/extdata/example.rda"))

> dim(sc_data)
[1] 22898   515

> table(sc_celltype)
sc_celltype
  Bcell Myeloid Stromal   Tcell   Tumor 
     83      38      23      54     317 

[2] create miRTalk object using single-cell transcriptomics data

> obj <- create_miRTalk(sc_data = sc_data,
                        sc_celltype = sc_celltype,
                        species = "Human",
                        condition = c(rep("Control", 250), rep("Disease", 265)),
                        evbiog = evbiog,
                        risc = risc,
                        ritac = ritac)
Warning: The following features are not present in the object: AGO2, not searching for symbol synonyms

[3] Find highly variable target genes with DEGs and HVGs with find_hvtg

> obj <- find_hvtg(object = obj)

[4] Find expressed miRNAs among all cells and generate background distribution for permutation test with find_miRNA.

> obj <- find_miRNA(object = obj,
                    mir_info = mir_info,
                    mir2tar = mir2tar)

[5] Infer cell-cell communication mediated by EV-derived miRNAs from senders to receivers under different conditions

> obj <- find_miRTalk(obj, if_doParallel = F)
[Control] 
[++++++++++++++++++++++++++++++] Finished:100% time:00:03:02
[Disease] 
[++++++++++++++++++++++++++++++] Finished:100% time:00:02:16

> obj_cci <- obj@cci
> nrow(obj_cci[obj_cci$condition == "Control",])
[1] 301

> nrow(obj_cci[obj_cci$condition == "Disease",])
[1] 194

Note: when there are multiple conditions in the obj, please set the parameter condition for the subsequent analysis and visulization since the default is to combine all miRNA-target interactions under different conditions for visulization

# condition = "Control" or condition = "Disease"
> plot_miRTalk_chord(object, condition = NULL, ...) 
> plot_miRTalk_circle(object, condition = NULL, ...)
> plot_miRTalk_circle_simple(object, condition = NULL, ...)
> plot_miRTalk_sankey(object, condition = NULL, ...)
> plot_miRTalk_heatmap(object, condition = NULL, ...)
> plot_miR_heatmap(object, condition = NULL, ...)
> plot_target_heatmap(object, condition = NULL, ...)
> plot_miR_bubble(object, condition = NULL, ...)
> plot_miR2tar_chord(object, condition = NULL, ...)
> plot_miR2tar_circle(object, condition = NULL, ...)
> plot_miR2tar_heatmap(object, condition = NULL, ...)