Merge pull request #21 from jumpingrivers/mut-lookups

Mutation lookup tables

DESCRIPTION

@@ -9,26 +9,29 @@ Description: A pipeline for scanning a SARS-CoV-2 phylogeny for clades with outl
 Depends:
   R (>= 4.1.0)
 Imports: 
-  ape,
-  dplyr,
-  foreach,
-  ggiraph,
-  ggplot2,
-  ggtree,
-  glue,
-  htmlwidgets,
-  knitr,
-  lubridate,
-  mgcv,
-  mlesky,
-  phangorn (>= 2.9.0),
-  rlang,
-  Rmpi,
-  scales,
-  stats,
-  stringr, 
-  treedater,
-  utils
+    ape,
+    dplyr,
+    foreach,
+    ggiraph,
+    ggplot2,
+    ggtree,
+    glue,
+    htmlwidgets,
+    knitr,
+    lubridate,
+    mgcv,
+    mlesky,
+    phangorn (>= 2.9.0),
+    purrr,
+    readr,
+    rlang,
+    Rmpi,
+    scales,
+    stats,
+    stringr,
+    tibble,
+    treedater,
+    utils
 Suggests: 
     doMPI,
     svglite,

R/parse_sc0.R

@@ -0,0 +1,49 @@
+#' Extract lists of 'defining' and 'all' mutations for each cluster in the data.frame `sc0`
+#'
+#' @param   sc0   data.frame. Must contains columns \code{cluster_id}, \code{defining_mutations} and
+#'   \code{all_mutations}. The "mutation" columns contain a "|"-separated string of the mutations
+#'   present in any given cluster. Each row represents a cluster in the phylogeny.
+#'
+#' @return   List of lists. Each entry in the named list corresponds to a cluster in the phylogeny.
+#'   The inner lists all have entries "defining" and "all", character vectors defining the mutations
+#'   that are present in the cluster.
+
+get_mutation_list <- function(sc0) {
+  required_columns <- c("defining_mutations", "all_mutations", "cluster_id")
+  stopifnot(all(required_columns %in% colnames(sc0)))
+
+  cmuts <- lapply(seq_len(nrow(sc0)), function(i) {
+    list(
+      defining = strsplit(sc0$defining_mutations[i], split = "\\|")[[1]],
+      all = strsplit(sc0$all_mutations[i], split = "\\|")[[1]]
+    )
+  })
+  names(cmuts) <- sc0$cluster_id
+
+  cmuts
+}
+
+#' Extract data.frames containing the 'defining' and 'all' mutations for each cluster
+#'
+#' @inheritParams   get_mutation_list
+#'
+#' @return   List of two data.frames with names "all" and "defining". These contain all mutations-
+#'   and just the defining mutations for each cluster in the phylogeny. The data.frames have
+#'   identical structure with column names \code{cluster_id} and \code{mutation}.
+
+get_mutation_tables <- function(sc0) {
+  mutation_list <- get_mutation_list(sc0)
+  defining_mutations <- purrr::map_df(
+    mutation_list, ~ tibble::tibble(mutation = .x[["defining"]]),
+    .id = "cluster_id"
+  )
+  all_mutations <- purrr::map_df(
+    mutation_list, ~ tibble::tibble(mutation = .x[["all"]]),
+    .id = "cluster_id"
+  )
+
+  list(
+    defining = defining_mutations,
+    all = all_mutations
+  )
+}

R/treeview.R

@@ -47,13 +47,15 @@ treeview <- function(e0,
     e0 <- readRDS(e0)
   }
   sc0 <- e0$Y
-  cmuts <- lapply(seq_len(nrow(sc0)), function(i) {
-    list(
-      defining = strsplit(sc0$defining_mutations[i], split = "\\|")[[1]],
-      all = strsplit(sc0$all_mutations[i], split = "\\|")[[1]]
-    )
-  })
-  names(cmuts) <- sc0$cluster_id
+
+  # 'cmuts' is a list. Each element has entries "defining" and "all". There is one entry for each
+  # node (row) in sc0
+  cmuts <- get_mutation_list(sc0)
+
+  cmut_tables <- get_mutation_tables(sc0)
+  readr::write_csv(cmut_tables[["defining"]], file.path(output_dir, "defining_mutations.csv"))
+  readr::write_csv(cmut_tables[["all"]], file.path(output_dir, "all_mutations.csv"))
+
   tr1 <- e0$tre
 
   stopifnot(all(branch_cols %in% colnames(e0$Y)))

inst/WORDLIST

@@ -40,11 +40,14 @@ phangorn
 phylo
 phylodynamics
 polymorphism
+purrr
+readr
 Reformats
 repo
 rlang
 Rmpi
 RoxygenNote
+sc
 scannint
 sina
 stringr
@@ -56,6 +59,7 @@ SystemRequirements
 testthat
 tfpscanner
 Thr
+tibble
 tis
 tooltips
 treedata