vis_comp_net_meas_boxplots2: Analyze and visualize network measure changes for Original,...

In ConesaLab/MDM: Complete MDM Network Analysis for 16S rRNA Data

Description

Creates customized boxplot of network measure changes across 3 network types for all ranks from Phylum to Genus for a given environment. Single or combinations of measures and ranks may also be selected. The statistical significance can be shown using stars (p.signif) or actual p-value scores (p.format).

Usage

vis_comp_net_meas_boxplots2(df, met_name, meas_name_options = NULL, rank_level_options = NULL, p_label = c("p.signif", "p.format"), p_size = 2)

Arguments

df	Dataframe of all scores per measure, per network type, and per rank for all nodes, created from comp_by_deleting_random_knowns_t_v3 function
met_name	Name of environment
meas_name_options	Network measures to compare. If left blank, will compare all network measures (degree, betweenness, and closeness). Otherwise, can select a single measure or combination of measures.
rank_level_options	Classification ranks to compare. If left blank, will compare network measure changes for all ranks from Phylum to Genus. Otherwise, can select a single rank or combination of ranks
p_label	Label of statistical significance. Can choose between 2 options - p.signif (using stars to denote significance) or p.format (using numerical p-value to denote significance)
p_size	Size of statistical significance indicators. Default is 2.

Value

Returns boxplot ggplot object comparing network measure changes across network types for the indicated rank levels.

Examples

##---- Should be DIRECTLY executable !! ----
##-- ==>  Define data, use random,
##--	or do  help(data=index)  for the standard data sets.

## The function is currently defined as
function (df, met_name, meas_name_options = NULL, rank_level_options = NULL,
    p_label = c("p.signif", "p.format"), p_size = 2)
{
    df = df[df$rank_level != "Kingdom" & df$rank_level != "Species",
        ]
    if (is.null(meas_name_options) && is.null(rank_level_options)) {
        g1 <- ggplot(df, aes(type, data, color = type)) + facet_grid(measure ~
            rank_level, scales = "free") + geom_boxplot(outlier.shape = NA) +
            ylab("Centrality score") + ggpubr::stat_compare_means(label = p_label,
            size = p_size, comparisons = list(c("Original", "Without_unknown"),
                c("Without_unknown", "Bootstrap"), c("Bootstrap",
                  "Original"))) + theme(axis.title.x = element_blank(),
            axis.text.x = element_blank())
    }
    else if (is.null(meas_name_options) && !(is.null(rank_level_options))) {
        spec_df = df[df$rank_level %in% rank_level_options, ]
        g1 <- ggplot(spec_df, aes(type, data, color = type)) +
            facet_grid(measure ~ rank_level, scales = "free") +
            geom_boxplot(outlier.shape = NA) + theme(axis.title.x = element_blank(),
            axis.text.x = element_blank()) + ggpubr::stat_compare_means(label = p_label,
            size = p_size, comparisons = list(c("Original", "Without_unknown"),
                c("Without_unknown", "Bootstrap"), c("Bootstrap",
                  "Original")))
    }
    else if (!(is.null(meas_name_options)) && is.null(rank_level_options)) {
        spec_df = df[df$measure %in% meas_name_options, ]
        g1 <- ggplot(spec_df, aes(type, data, color = type)) +
            facet_grid(measure ~ rank_level, scales = "free") +
            geom_boxplot(outlier.shape = NA) + theme(axis.title.x = element_blank(),
            axis.text.x = element_blank()) + ggpubr::stat_compare_means(label = p_label,
            size = p_size, comparisons = list(c("Original", "Without_unknown"),
                c("Without_unknown", "Bootstrap"), c("Bootstrap",
                  "Original")))
    }
    else {
        spec_df = df[df$measure %in% meas_name_options & df$rank_level %in%
            rank_level_options, ]
        g1 <- ggplot(spec_df, aes(type, data, color = type)) +
            geom_boxplot(outlier.shape = NA)
        if (length(meas_name_options) > 1 && length(rank_level_options) >
            1) {
            g1 <- g1 + facet_grid(measure ~ rank_level, scales = "free") +
                ggpubr::stat_compare_means(label = p_label, size = p_size,
                  comparisons = list(c("Original", "Without_unknown"),
                    c("Without_unknown", "Bootstrap"), c("Bootstrap",
                      "Original"))) + theme(axis.title.x = element_blank(),
                axis.text.x = element_blank())
        }
        else if (length(meas_name_options) > 1 && length(rank_level_options) ==
            1) {
            g1 <- g1 + facet_wrap(~measure, scales = "free") +
                ggpubr::stat_compare_means(label = p_label, size = p_size,
                  comparisons = list(c("Original", "Without_unknown"),
                    c("Without_unknown", "Bootstrap"), c("Bootstrap",
                      "Original"))) + theme(axis.title.x = element_blank(),
                axis.text.x = element_blank())
        }
        else if (length(meas_name_options) == 1 && length(rank_level_options) >
            1) {
            g1 <- g1 + facet_wrap(~rank_level, scales = "free") +
                ggpubr::stat_compare_means(label = p_label, size = p_size,
                  comparisons = list(c("Original", "Without_unknown"),
                    c("Without_unknown", "Bootstrap"), c("Bootstrap",
                      "Original"))) + theme(axis.title.x = element_blank(),
                axis.text.x = element_blank())
        }
        else if (length(meas_name_options) == 1 && length(rank_level_options) ==
            1) {
            g1 <- g1 + ylab(meas_name_options) + ggtitle(rank_level_options) +
                ggpubr::stat_compare_means(label = p_label, comparisons = list(c("Original",
                  "Without_unknown"), c("Without_unknown", "Bootstrap"),
                  c("Bootstrap", "Original"))) + theme(axis.title.x = element_blank(),
                axis.text.x = element_blank())
        }
    }
    g1 <- g1 + ggtitle(paste(met_name, "Centrality Score Comparison",
        sep = " ")) + ylab("Centrality score") + theme(plot.title = element_text(hjust = 0.5))
    return(g1)
  }

ConesaLab/MDM documentation built on Aug. 1, 2020, 11:47 a.m.