R/debugging.R

In EngeLab/CIMseq: CIMseq is a software suite for analysis of scRNAseq data that has been produced using the cellular interactions by multiplet sequencing method

#'@include All-classes.R
NULL

#' plotSwarmPosition
#'
#' Subtitle
#'
#' Plots the swarm position in a spSwarm object for each reported iteration
#' (specified by reportRate) when report equals TRUE.
#'
#' @name plotSwarmPosition
#' @rdname plotSwarmPosition
#' @aliases plotSwarmPosition
#' @param swarm A CIMseqSwarm object.
#' @param multiplet The name of the multiplet to plot the swarm positions for.
#' @param interval The iterations to be plotted if all recorded iteration
#'    intervals should not be plotted.
#' @param ... additional arguments to pass on.
#' @return The plotSwarmPosition function returns an object of class spCounts.
#' @author Jason T. Serviss
#' @keywords plotSwarmPosition
#' @examples
#'
#' #no test data where swarm was tracked with report arg.
NULL

#' @rdname plotSwarmPosition
#' @export

setGeneric("plotSwarmPosition", function(
  swarm, ...
){
  standardGeneric("plotSwarmPosition")
})

#' @rdname plotSwarmPosition
#' @export
#' @import ggplot2
#' @importFrom dplyr "%>%" filter mutate bind_rows
#' @importFrom tibble as_tibble
#' @importFrom tidyr gather
#' @importFrom ggthemes theme_few
#' @importFrom readr parse_factor
#' @importFrom rlang .data

setMethod("plotSwarmPosition", "CIMseqSwarm", function(
    swarm,
    multiplet,
    interval = NULL,
    ...
){
  #get the reported statistics and subset only those multiplets of interest
  stats <- getData(swarm, "stats")
  classes <- colnames(getData(swarm, "fractions"))
  names(stats) <- rownames(getData(swarm, "fractions"))
  stats <- stats[names(stats) %in% multiplet]
  
  #reformat all the matrixes as tibbles and add the appropriate names
  ll <- stats %>%
    lapply(.data, `[[`, 4) %>%
    lapply(.data %>%
    lapply(.data %>%
    as_tibble() %>%
    setNames(paste("v", 1:ncol(.data), sep = "")) %>%
    mutate(class = .data$classes)
  ))
  
  ll <- lapply(ll, setNames, paste(1:100, sep = ""))
  
  position <- lapply(1:length(ll), function(x) {
    bind_rows(ll[[x]], .id = "iteration")
  })
  
  p <- position %>%
    setNames(names(stats)) %>%
    bind_rows(.data, .id = "multiplet") %>%
    gather(
      "swarmMember", "position", -.data$iteration, -.data$multiplet,
      -.data$class
    ) %>%
    mutate(
      iteration = parse_factor(
        .data$iteration,
        levels = unique(.data$iteration)
      )
    ) %>%
    mutate(class = factor(.data$class)) %>%
    mutate(position = log2(.data$position + 1)) %>%
    filter(.data$iteration %in% interval) %>%
    as_tibble() %>%
    ggplot(aes(position)) +
      geom_density(aes(fill = class, colour = class), alpha = 0.5) +
      facet_grid(multiplet ~ iteration, scale = "free") +
      scale_fill_manual(values = col40()) +
      scale_colour_manual(values = col40()) +
      theme_few() +
      theme(
        legend.position = "top",
        strip.text.y = element_text(angle = 0)
      ) +
      guides(
        fill = guide_legend(
          ncol = 8, title = "Class", title.position = "top", title.hjust = 0.5
        ),
        colour = FALSE
      ) +
      labs(
        x = "Swarm position", y = "Density")
  p
  return(p)
})