score_gsva_likelihoods: Score the results from simple_gsva().

In elsayed-lab/hpgltools: A pile of (hopefully) useful R functions

Score the results from simple_gsva().

Description

Yeah, this is a bit meta, but the scores from gsva seem a bit meaningless to me, so I decided to look at the distribution of observed scores in some of my data; I quickly realized that they follow a nicely normal distribution. Therefore, I thought to calculate some scores of gsva() using that information.

Usage

score_gsva_likelihoods(
  gsva_result,
  score = NULL,
  category = NULL,
  factor = NULL,
  sample = NULL,
  factor_column = "condition",
  method = "mean",
  label_size = NULL,
  col_margin = 6,
  row_margin = 12,
  cutoff = 0.95
)

Arguments

gsva_result	Input result from simple_gsva()
score	What type of scoring to perform, against a value, column, row?
category	What category to use as baseline?
factor	Which experimental factor to compare against?
sample	Which sample to compare against?
factor_column	When comparing against an experimental factor, which design column to use to find it?
method	mean or median when when bringing together values?
label_size	By default, enlarge the labels to readable at the cost of losing some.
col_margin	Attempt to make heatmaps fit better on the screen with this and...
row_margin	this parameter
cutoff	Highlight only the categories deemed more significant than this.

Details

The nicest thing in this, I think, is that it provides its scoring metric(s) according to a few different possibilities, including: * the mean of samples found in an experimental factor * All provided scores against the distribution of observed scores as z-scores. * A single score against all scores. * Rows (gene sets) against the set of all gene sets.

Value

The scores according to the provided category, factor, sample, or score(s).

See Also

[simple_gsva()]

elsayed-lab/hpgltools documentation built on May 9, 2024, 5:02 a.m.