R/plot_API.R

In APackOfTheClones: Visualization of Clonal Expansion for Single Cell Immune Profiles

# main apotc plot initializer
create_initial_apotc_plot <- function(
	apotc_obj, res, linetype, alpha, detail = TRUE
) {
 
  if (!detail) {
    plt_df <- make_undetailed_df(apotc_obj)
  } else {
    plt_df <- get_plottable_df_with_color(apotc_obj)
  }

	plot_clusters(
    clusters = plt_df,
    n = res,
    linetype = linetype,
    alpha = alpha
  )
}

plot_clusters <- function(
	clusters, # is a df
	n = 360,
	linetype = "blank",
	alpha = 1
) {
	clusters %>%
		ggplot2::ggplot() +
		ggforce::geom_circle(
			apotc_aes_string(
				x0 = "x",
				y0 = "y",
				r = "r",
				fill = "color"
			),
			n = n,
			linetype = linetype,
			alpha = alpha
		) +
		ggplot2::coord_fixed() +
		ggplot2::scale_fill_identity()
}

get_plottable_df_with_color <- function(apotc_data) {
    extract_and_add_colors(
        apotc_obj = apotc_data,
        plot_df = df_full_join(get_clusterlists(apotc_data))
    )
}

# full join a list of clusterlists vectors into a dataframe with
# generated labels.
df_full_join <- function(clstr_list, detail = TRUE) {

    df <- data.frame(
        "label" = character(0),
        "x" = numeric(0),
        "y" = numeric(0),
        "r" = numeric(0)
    ) %>%
        update_clusterlist_df(get_first_nonempty(clstr_list))

    if (contains_clonotypes(get_first_nonempty(clstr_list))) {
        cols_to_join_by <- dplyr::join_by("label", "x", "y", "r", "clonotype")
    } else {
        cols_to_join_by <- dplyr::join_by("label", "x", "y", "r")
    }

    seurat_cluster_index <- 0

    for (i in seq_along(clstr_list)) {

        seurat_cluster_index <- seurat_cluster_index + 1
        if (is_empty(clstr_list[[i]])) next

        df <- df %>% dplyr::full_join(
            convert_to_dataframe(clstr_list[[i]], seurat_cluster_index - 1), # zero indexed
            by = cols_to_join_by
        )
    }

    df
}

make_undetailed_df <- function(apotc_obj) {

	df <- data.frame(
		label = character(0),
    x = numeric(0),
    y = numeric(0),
    r = numeric(0),
    color = character(0),
    clonotype = character(0)
	)

	for (el in enumerate(get_clusterlists(apotc_obj))) {
    if (is_empty(val1(el))) next
    df <- df %>% dplyr::full_join(
      convert_to_dataframe(val1(el), ind(el) - 1, detail = FALSE) %>%
        dplyr::mutate(
          "clonotype" = "df_generated_with_detail=FALSE",
          "color" = get_cluster_colors(apotc_obj)[[ind(el)]]
        ),
      by = dplyr::join_by("label", "x", "y", "r", "clonotype", "color")
    )
	}

  df
}

# check if plot was made with detail
is_undetailed <- function(apotc_ggplot) {
  clonotypes <- get_ggplot_data(apotc_ggplot)$clonotype
  is.null(clonotypes) || (clonotypes[1] == "df_generated_with_detail=FALSE")
}

add_default_theme <- function(plt, reduction) {
	label_hashmap <- hash::hash(
		  c("umap", "tsne", "pca"), c("UMAP", "tSNE", "PC")
	)
	label <- label_hashmap[[reduction]]
  if (is.null(label)) label <- reduction

	plt +
      ggplot2::theme_classic() +
      ggplot2::xlab(paste(label, 1, sep = "_")) +
      ggplot2::ylab(paste(label, 2, sep = "_"))
}

get_retain_scale_dims <- function(
  seurat_obj, reduction, ball_pack_plt, plot_dims
) {

  reduction_dims <- get_plot_dims(
    Seurat::DimPlot(object = seurat_obj, reduction = reduction)
  )
  reduction_xr <- get_xr(reduction_dims)
  reduction_yr <- get_yr(reduction_dims)

  ball_pack_xr <- get_xr(plot_dims)
  ball_pack_yr <- get_yr(plot_dims)

  # set new ranges
  min_xr <- min(ball_pack_xr[1], reduction_xr[1])
  max_xr <- max(ball_pack_xr[2], reduction_xr[2])

  min_yr <- min(ball_pack_yr[1], reduction_yr[1])
  max_yr <- max(ball_pack_yr[2], reduction_yr[2])

  # return dims in same output format as get_plot_dims
  list("xr" = c(min_xr, max_xr), "yr" = c(min_yr, max_yr))

}

check_is_apotc_ggplot <- function(x) {
  if (!isApotcGGPlot(x)) {
    stop(call. = FALSE, "not an output of `APOTCPlot()` or `vizAPOTC()`")
  }
}