Compute Z-Score and P-Value for Node Pagerank — compute_network

This function calculates the z-score and p-value for a node's pagerank by comparing it to a set of permutation values. It returns a data frame containing the z-score (beta) and p-value (p_value).

Usage

compute_network_enrichment(node)

Arguments

node: Either a named list / one-row data frame with pagerank and permutation values whose names start with "perm" (e.g. perm.1, perm.2), or a data frame with one row per node in that layout, such as the pagerank_data slot of a run_tkoi() result with keep_permutations = TRUE. A list of equal-length vectors is treated as such a data frame. node needs a pagerank value and at least two perm* values, or an error is raised.

Value

A data frame with one row per node and columns:

beta: The computed z-score.
p_value: The one-tailed p-value derived from the z-score.

Details

The function calculates the z-score as: $$z = \frac{\text{pagerank} - \text{mean}(\text{perm_values})}{\text{sd}(\text{perm_values})}$$ (NaN when the permutation values have no spread, as in run_tkoi), and the p-value is calculated as the survival function of the z-score: $$p = 1 - \Phi(z)$$ where $\Phi$ is the cumulative distribution function of the standard normal distribution.

Examples

compute_network_enrichment(list(pagerank = 0.3, perm.1 = 0.1, perm.2 = 0.2, perm.3 = 0.15))
#>   beta     p_value
#> 1    3 0.001349898

nodes = data.frame(pagerank = c(0.3, 0.1), perm.1 = c(0.1, 0.1), perm.2 = c(0.2, 0.12))
compute_network_enrichment(nodes)
#>         beta    p_value
#> 1  2.1213203 0.01694743
#> 2 -0.7071068 0.76024994