meta_plot: Candidate Drivers Search Plot

In montilab/CaDrA: Candidate Driver Analysis

Candidate Drivers Search Plot

Description

By utilizing the top N results obtained from candidate_search, we can find the best meta-feature among the top N searches using topn_best. meta_plot is then used to produce graphics including a tile plot for the top meta-features that associated with a molecular phenotype of interest (e.g. input_score), the KS enrichment plot of the meta-features, and lastly, a density diagram of the distribution of the observed input scores sorted from largest to smallest at the top.

Usage

meta_plot(topn_best_list, input_score_label = NULL, plot_title = NULL)

Arguments

topn_best_list	a list of objects returned from candidate_search corresponding to the search of top N features given by top_N value. The topn_best_list contains the best meta-feature matrix, its corresponding best score, its observed input scores, rank of the best features based on their scores, marginal best scores, and cumulative best scores.
input_score_label	a label that references to the input_score variable that was used to compute the top N best features. Default is NULL.
plot_title	a title to the plot. Default is NULL.

Value

3 plots stacked on top of each other: 1. a density diagram of observed input scores sorted from highest to lowest 2. a tile plot of the top features within the meta-feature set 3. a KS enrichment plot of the meta-feature set (this correspond to the logical OR of the features)

Examples

# Load pre-computed Top-N list generated for sim_FS dataset
data(topn_list)

# With the results obtained from top-N evaluation,
# We can find the combination of features that gives the best score in
# top N searches
topn_best_meta <- topn_best(topn_list = topn_list)

# Now we can plot this set of best meta-feature
meta_plot(topn_best_list = topn_best_meta)

montilab/CaDrA documentation built on Aug. 22, 2024, 11:55 p.m.