Merge master branch so pkgdown will be up to date

.Rbuildignore

@@ -18,4 +18,3 @@
 ^old\.README\.md$
 ^readme_images$
 ^scenic_bash$
-^_pkgdown\.yml$

.github/workflows/bioc_branch.yml

@@ -0,0 +1,4 @@
+name: Update Bioconductor package files
+
+on:
+  workflow_dispatch:

.github/workflows/r-build-check.yml

@@ -0,0 +1,12 @@
+name: r-build-check
+on:
+    push:
+        branches: ['v0.2.2_development']
+    pull_request:
+        branches: ['master', 'v0.2.2']
+    workflow_dispatch:
+
+jobs:
+    build-check:
+        uses: FertigLab/actions/.github/workflows/r-build-check.yml@v1
+        secrets: inherit

.gitignore

@@ -10,5 +10,3 @@ R/.Rhistory
 scenic/*
 
 scratch/
-
-docs

NAMESPACE

@@ -0,0 +1,55 @@
+# Generated by roxygen2: do not edit by hand
+
+export(add_rl_column)
+export(build_domino)
+export(circos_ligand_receptor)
+export(cor_heatmap)
+export(cor_scatter)
+export(count_linkage)
+export(create_domino)
+export(create_regulon_list_scenic)
+export(create_rl_map_cellphonedb)
+export(dom_clusters)
+export(dom_correlations)
+export(dom_counts)
+export(dom_database)
+export(dom_de)
+export(dom_info)
+export(dom_linkages)
+export(dom_network_items)
+export(dom_signaling)
+export(dom_tf_activation)
+export(dom_zscores)
+export(feat_heatmap)
+export(gene_network)
+export(incoming_signaling_heatmap)
+export(mean_ligand_expression)
+export(plot_differential_linkages)
+export(rename_clusters)
+export(signaling_heatmap)
+export(signaling_network)
+export(summarize_linkages)
+export(test_differential_linkages)
+exportClasses(domino)
+exportClasses(linkage_summary)
+import(ComplexHeatmap)
+import(biomaRt)
+import(circlize)
+import(grDevices)
+import(grid)
+import(methods)
+import(plyr)
+import(stats)
+importClassesFrom(Matrix,dgCMatrix)
+importFrom(Matrix,rowSums)
+importFrom(ggpubr,ggscatter)
+importFrom(igraph,E)
+importFrom(igraph,V)
+importFrom(igraph,graph)
+importFrom(igraph,layout_in_circle)
+importFrom(igraph,layout_on_sphere)
+importFrom(igraph,layout_randomly)
+importFrom(igraph,layout_with_fr)
+importFrom(igraph,layout_with_kk)
+importFrom(igraph,simplify)
+importFrom(utils,read.csv)

NEWS.md

@@ -0,0 +1,31 @@
+# domino2 v0.2.2
+
+## Linkage functions
+- Addition of new class to summarize linkages in objects
+- Addition of helper functions to count linkages and compare between objects
+- Plotting function for differential linkages
+
+## Package structure
+- Adjustments made to meet BioConductor standards
+
+# domino2 v0.2.1
+
+## Updates to domino object construction
+- Uniform formats for inputs of receptor-ligand interaction databases, transcription factor activity features, and regulon gene lists for operability with alternative databases and transcription factor activation inference methods
+- Helper functions for reformatting pySCENIC outputs and CellPhoneDB database files to domino-readable uniform formats
+- Assessment of transcription factor linkage with receptors that function as a heteromeric complex based on correlation between transcription factor activity and all receptor component genes
+- Assessment of complex ligand expression as the mean of component gene expression for plotting functions
+- Minimum threshold for the percentage of cells in a cluster expressing a receptor gene for the receptor to be called active within the cluster
+- Additional linkage slots for active receptors in each cluster, transcription factor-receptor linkages for each cluster, and incoming ligands for active receptors on each cluster
+
+## Plotting functions
+- Chord plot of ligand expression targeting a specified receptor where chord widths correspond to the quantity of ligand expression by each cell cluster
+- Signaling networks showing only outgoing signaling from specified cell clusters
+- Gene networks between two cell clusters
+
+## Bugfixes
+- Added host option for gene ortholog conversions using `{biomaRt}` for access to maintained mirrors
+- Transcription factor-target linkages are now properly stored so that receptors in a transcription factor's regulon are excluded from linkage
+- Ligand nodes sizes in gene networks correspond to quantity of ligand expression
+- `create_domino()` can be run without providing a regulon list
+- References to the host GitHub repository have been updated to [Elisseeff-Lab](https://github.com/Elisseeff-Lab/domino)

R/class_definitions.R

@@ -0,0 +1,121 @@
+#' @import methods
+#' @importClassesFrom Matrix dgCMatrix
+#'
+NULL
+#' The domino Class
+#'
+#' The domino class contains all information necessary to calculate receptor-ligand
+#' signaling. It contains z-scored expression, cell cluster labels, feature values,
+#' and a referenced receptor-ligand database formatted as a receptor-ligand map.
+#' Calculated intermediate values are also stored.
+#'
+#' @slot db_info List of data sets from lr database.
+#' @slot counts Raw count gene expression data
+#' @slot z_scores Matrix of z-scored expression data with cells as columns
+#' @slot clusters Named factor with cluster identity of each cell
+#' @slot features Matrix of features to correlate receptor-ligand expression with. Cells are columns and features are rows.
+#' @slot cor Correlation matrix of receptor expression to features.
+#' @slot linkages List of lists containing info linking cluster->tf->rec->lig
+#' @slot clust_de Data frame containing differential expression results for features by cluster.
+#' @slot misc List of miscellaneous info pertaining to run parameters etc.
+#' @slot cl_signaling_matrices Incoming signaling matrix for each cluster
+#' @slot signaling Signaling matrix between all clusters.
+#' @name domino-class
+#' @rdname domino-class
+#' @exportClass domino
+#'
+domino <- methods::setClass(
+  Class = "domino",
+  slots = c(
+    db_info = "list",
+    z_scores = "matrix",
+    counts = "dgCMatrix",
+    clusters = "factor",
+    features = "matrix",
+    cor = "matrix",
+    linkages = "list",
+    clust_de = "matrix",
+    misc = "list",
+    cl_signaling_matrices = "list",
+    signaling = "matrix"
+  ),
+  prototype = list(
+    misc = list("build" = FALSE)
+  )
+)
+#' The domino linkage summary class
+#'
+#' The linkage summary class contains linkages established in multiple domino
+#' objects through gene regulatory network inference and reference to receptor-
+#' ligand data bases. A data frame summarizing meta features that describe the
+#' domino objects compared in the linkage summary facilitates comparisons of
+#' established linkages and differential signaling interactions across categorical
+#' sample covariates.
+#'
+#' @slot subject_names unique names for each domino result included in the summary
+#' @slot subject_meta data.frame with each row describing one subject and columns describing features of the subjects by which to draw comparisons of signaling networks
+#' @slot subject_linkages nested list of linkages inferred for each subject. Lists are stored in a heirarchical structure of subject-cluster-linkage where linkages include transcription factors (tfs), linkages between transcription factors and receptors (tfs_rec), active receptors (rec), possible receptor-ligand interactions (rec_lig), and incoming ligands (incoming_lig)
+#' @name linkage_summary-class
+#' @rdname linkage_summary-class
+#' @exportClass linkage_summary
+#'
+linkage_summary <- setClass(
+  Class = "linkage_summary",
+  slots = c(
+    subject_names = "factor",
+    subject_meta = "data.frame",
+    subject_linkages = "list"
+  )
+)
+
+#' Print domino object
+#'
+#' Prints a summary of a domino object
+#'
+#' @param x Domino object
+#' @return a printed description of the number of cell clusters in the object
+#' @keywords internal
+#' @examples
+#' print(domino2:::pbmc_dom_built_tiny)
+#' 
+setMethod("print", "domino", function(x, ...) {
+  if (x@misc$build) {
+    message(
+      "A domino object of ", length(x@clusters), " cells
+                Contains signaling between",
+      length(levels(x@clusters)), "clusters
+                Built with a maximum of", as.integer(x@misc$build_vars["max_tf_per_clust"]),
+      "TFs per cluster
+                and a maximum of", as.integer(x@misc$build_vars["max_rec_per_tf"]),
+      "receptors per TF\n"
+    )
+  } else {
+    message(c("A domino object of ", length(x@clusters), " cells\n", "A signaling network has not been built\n"),
+      sep = ""
+    )
+  }
+})
+#' Show domino object information
+#'
+#' Shows content overview of domino object
+#'
+#' @param object Domino object
+#' @return a printed description of the number of cells in a domino object and its build status
+#' @keywords internal
+#' @examples
+#' domino2:::pbmc_dom_built_tiny
+#' 
+#' show(domino2:::pbmc_dom_built_tiny)
+#' 
+setMethod("show", "domino", function(object) {
+  if (object@misc$build) {
+    cat(c(
+      "A domino object of ", length(object@clusters), " cells\n", "Built with signaling between ",
+      length(levels(object@clusters)), " clusters\n"
+    ), sep = "")
+  } else {
+    cat(c("A domino object of ", length(object@clusters), " cells\n", "A signaling network has not been built\n"),
+      sep = ""
+    )
+  }
+})

R/convenience_fxns.R

@@ -0,0 +1,99 @@
+#' @import plyr
+#' @import methods
+#'
+NULL
+
+#' Renames clusters in a domino object
+#'
+#' This function reads in a receptor ligand signaling database, cell level
+#' features of some kind (ie. output from pySCENIC), z-scored single cell data,
+#' and cluster id for single cell data, calculates a correlation matrix between
+#' receptors and other features (this is transcription factor module scores if
+#' using pySCENIC), and finds features enriched by cluster. It will return a
+#' domino object prepared for [build_domino()], which will calculate a signaling
+#' network.
+#'
+#' @param dom Domino object to rename clusters in
+#' @param clust_conv Named vector of conversions from old to new clusters. Values are taken as new clusters IDs and names as old cluster IDs.
+#' @return A domino object with clusters renamed in all applicable slots.
+#' @keywords internal
+#' @export
+#' @examples 
+#' new_clust <- c("CD8_T_cell" = "CD8+ T Cells",
+#'  "CD14_monocyte" = "CD14+ Monocytes", "B_cell" = "B Cells")
+#' pbmc_dom_built_tiny <- rename_clusters(domino2:::pbmc_dom_built_tiny, new_clust)
+#'
+rename_clusters <- function(dom, clust_conv) {
+  if (is.null(dom@clusters)) {
+    stop("There are no clusters in this domino object")
+  }
+  if (dom@misc$create) {
+    dom@clusters <- revalue(dom@clusters, clust_conv)
+    colnames(dom@clust_de) <- clust_conv
+    names(colnames(dom@clust_de)) <- c()
+    colnames(dom@misc$cl_rec_percent) <- clust_conv
+  }
+  if (dom@misc$build) {
+    names(dom@linkages$clust_tf) <- clust_conv
+    names(dom@linkages$clust_rec) <- clust_conv
+    names(dom@linkages$clust_incoming_lig) <- clust_conv
+    names(dom@linkages$clust_tf_rec) <- clust_conv
+    colnames(dom@signaling) <- paste0("L_", clust_conv)
+    rownames(dom@signaling) <- paste0("R_", clust_conv)
+    names(dom@cl_signaling_matrices) <- clust_conv
+    for (cl in clust_conv) {
+      colnames(dom@cl_signaling_matrices[[cl]]) <- paste0("L_", clust_conv)
+    }
+  }
+  return(dom)
+}
+
+#' Convert Genes Using Table
+#'
+#' Takes a vector of gene inputs and a conversion table  and returns a 
+#' converted gene table
+#'
+#' @param genes The genes to convert.
+#' @param from  Gene symbol type of the input (ENSG, ENSMUSG, HGNC, MGI)
+#' @param to    Desired gene symbol type for the output (HGNC, MGI)
+#' @param conversion_table A data.frame with column names corresponding to gene symbol types (mm.ens, hs.ens, mgi, hgnc)
+#' and rows corresponding to the gene symbols themselves
+#' @return Data frame of genes with original and corresponding converted symbols
+#' @keywords internal
+#'
+table_convert_genes <- function(genes, from, to, conversion_table) {
+  # Check inputs:
+  stopifnot(`Genes must be a vector of characters` = (is(genes, "character") & is(genes, "vector")))
+  stopifnot(`From must be one of ENSMUSG, ENSG, MGI, or HGNC` = from %in% c(
+    "ENSMUSG", "ENSG", "MGI",
+    "HGNC"
+  ))
+  stopifnot(`To must be one of MGI or HGNC` = to %in% c("MGI", "HGNC"))
+  stopifnot(`Conversion table must be provided with at least two of column names mm.ens, hs.ens, mgi and/or hgnc` = (is(
+    conversion_table,
+    "data.frame"
+  ) & length(which(colnames(conversion_table) %in% c(
+    "mm.ens", "hs.ens", "mgi",
+    "hgnc"
+  ))) > 1))
+  if (from == "ENSMUSG") {
+    col1 <- conversion_table$mm.ens
+  }
+  if (from == "ENSG") {
+    col1 <- conversion_table$hs.ens
+  }
+  if (from == "MGI") {
+    col1 <- conversion_table$mgi
+  }
+  if (from == "HGNC") {
+    col1 <- conversion_table$hgnc
+  }
+  if (to == "MGI") {
+    col2 <- conversion_table$mgi
+  }
+  if (to == "HGNC") {
+    col2 <- conversion_table$hgnc
+  }
+  genesV2 <- cbind(col1[which(col1 %in% genes)], col2[which(col1 %in% genes)])
+  return(genesV2)
+}