Skip to contents

Create Starlng Shiny app from Monocle object

Source: R/shiny.R
starlng_write_app_monocle.Rd

Runs the entire Starlng pipeline on a Monocle object and generates a folder with all necessary files to run a Shiny app.

Usage

starlng_write_app_monocle(
  folder_path,
  monocle_object,
  app_title_name = "",
  learn_graph_parameters = list(nodes_per_log10_cells = 30, learn_graph_controls =
    list(eps = 1e-05, maxiter = 100)),
  gene_filtering_function = function(info_gene_df) {
     rownames(info_gene_df %>%
    dplyr::filter(.data$morans_I > 0.1, .data$q_value < 0.05))
 },
  clustering_parameters = list(n_neighbours = seq(from = 5, to = 50, by = 5), graph_type
    = "snn", prune_value = -1, resolutions = list(RBConfigurationVertexPartition =
    seq(from = 0.1, to = 2, by = 0.1), RBERVertexPartition = NULL,
    ModularityVertexPartition = NULL), number_iterations = 5, number_repetitions = 100),
  ecc_threshold = 0.9,
  freq_threshold = 30,
  enrichment_organism = "hsapiens",
  save_entire_monocle = TRUE,
  discrete_colours = list(),
  continuous_colours = list(),
  max_n_colors = 40,
  verbose = FALSE,
  compression_level = 7,
  chunk_size = 100,
  nthreads = 1
)

Arguments

folder_path

The path of the folder where the app will be saved.

monocle_object

The monocle3 object that will be used as a starting point.

app_title_name

The title of the shiny app.

learn_graph_parameters

A list of parameters that will be passed to the custom_learn_graph function. Possible options:

  • nodes_per_log10_cells - the number of nodes per log10 cells. It is used to influence the level of detail in the inferred trajectory.

  • learn_graph_controls - a list of parameters that will be passed to the learn_graph function of the monocle3 package. These parameters are usually used to influence the convergence of the algorithm or the pruning of the resulting trajectory.

gene_filtering_function

A function that will be used to filter the genes that will be used in the clustering analysis. The function should take only one parameter, which is a data frame. The data frame will contain the following columns:

  • morans_I - the Moran's I value of the gene.

  • q_value - the adjusted p-value of the gene being spatially autocorrelated.

  • average_expression - the average expression of the gene.

  • n_expressed_cells - the number of cells that express the gene.

  • average_expression_nonzero - the average expression of the gene in cells that express it.

  • percent_expressed_cells - the percentage of cells that express the gene. The function should return a list of gene names that will be considered to be the result of the filtering.

clustering_parameters

A list of parameters that will be passed to the clustering_pipeline function. Consult the documentation of the function for further details.

ecc_threshold

The threshold applied on the Element-Centric Consistency (ECC) score to determine if a number of cluster is stable or not. By default, it is set to 0.9.

freq_threshold

The threshold applied on the number of times a number of clusters is obtained after running the clustering_pipeline function. By default, it is set to 30.

enrichment_organism

The organism used for the enrichment analysis.

save_entire_monocle

If TRUE, saves the monocle object in the app folder. This object could be used to perform additional changes to the app or to continue the downstream analysis. If memory is a concern, set this parameter to FALSE. Defaults to TRUE.

discrete_colours

A list of colours that will be used to colour the groups from the discrete metadata. The list should consist in an association between the number of groups and a colour palette. For the missing number of groups, the qualpalr package will be used. Defaults to an empty list.

continuous_colours

A list of colours that will be used to colour the continuous metadata and the gene expression. The list should consist in an association between the name of the colour palette and a list of colours. To this list the following palettes will be automatically added: viridis, plasam, white-red and blue-red. Defaults to an empty list.

max_n_colors

The maximum number of colours used for discrete groups. Defaults to 40.

verbose

If TRUE, prints intermediate progress messages of the pipeline. Defaults to FALSE.

compression_level

The compression level used to save the expression matrix in the HDF5 format. Defaults to 7.

chunk_size

The chunk size used to store the expression matrix in the HDF5 format. It is recommended to use values between 10 and 1000, as they might improve the speed of reading the rows from the file. Defaults to 100.

nthreads

The number of threads used to perform the calculations. This parameter will be automatically passed to the graph_test function that will calculate the Moran's I test. For the clustering pipeline, if no parallel backend is already registered, the function will generate a PSOCK cluster with the number of threads specified. Defaults to 1.

Value

The function does not return anything, but it creates a folder which contains all necessary files to run the Starlng Shiny app.

  • app.R - the main file of the Shiny app.

  • objects - a folder that contains the following files:

  • monocle_object.qs2 - the monocle object that was used to generate the app.

  • recommended_pseudotime.qs2 - the recommended pseudotime values.

  • genes_info.csv - a table with the gene information.

  • trajectory_ggplot.qs2 - the ggplot object of the trajectory.

  • metadata.qs2 - the metadata object.

  • diet_monocle_object.qs2 - the monocle object with the diet information.

  • expression.h5 - the expression matrix in the HDF5 format.

  • colours.qs2 - the colours object.