Generate data files required for shiny app. Five files will be generated, namely (i) the shinycell config prefix_conf.rds, (ii) the gene mapping object config prefix_gene.rds, (iii) the single-cell gene expression prefix_gexpr.h5, (iv) the single-cell metadata prefix_meta.rds and (v) the defaults for the Shiny app prefix_def.rds. A prefix is specified in each file to allow for the use of multiple single-cell datasets in one Shiny app. Note that both make_file and make_code functions are ran when running the wrapper function make_app.

make_file(
  obj,
  scConf,
  gex.assay = NA,
  gex.slot = c("data", "scale.data", "counts"),
  gene.mapping = FALSE,
  shiny.prefix = "sc1",
  shiny.dir = "shinyApp/",
  default.gene1 = NA,
  default.gene2 = NA,
  default.multigene = NA,
  default.dimred = NA,
  chunkSize = 500,
  mar = FALSE
)

Arguments

obj

input single-cell object for Seurat (v3+) / SingleCellExperiment data or input file path for h5ad / loom files

scConf

shinycell config data.table

gex.assay

assay in single-cell data object to use for plotting gene expression, which must match one of the following:

  • Seurat objects: "RNA" or "integrated" assay, default is "RNA"

  • SCE objects: "logcounts" or "normcounts" or "counts", default is "logcounts"

  • h5ad files: "X" or any assay in "layers", default is "X"

  • loom files: "matrix" or any assay in "layers", default is "matrix"

gex.slot

slot in single-cell assay to plot. This is only used for Seurat objects (v3+). Default is to use the "data" slot

gene.mapping

specifies whether to convert human / mouse Ensembl gene IDs (e.g. ENSG000xxx / ENSMUSG000xxx) into "user-friendly" gene symbols. Set this to TRUE if you are using Ensembl gene IDs. Default is FALSE which is not to perform any conversion. Alternatively, users can supply a named vector where names(gene.mapping) correspond to the actual gene identifiers in the gene expression matrix and gene.mapping correspond to new identifiers to map to

shiny.prefix

specify file prefix

shiny.dir

specify directory to create the shiny app in

default.gene1

specify primary default gene to show

default.gene2

specify secondary default gene to show

default.multigene

character vector specifying default genes to show in bubbleplot / heatmap

default.dimred

character vector specifying the two default dimension reductions. Default is to use UMAP if not TSNE embeddings

chunkSize

number of genes written to h5file at any one time. Lower this number to reduce memory consumption. Should not be less than 10

mar

Logical indicating if marker genes are used

Value

data files required for shiny app

Author

John F. Ouyang

Roy Francis