make_correlation_network: Construct igraph based on pairwise spearman correlations

In pmtran5884/Glycancc: Community Clustering of Anti-Carbohydrate Antibody data

Construct igraph based on pairwise spearman correlations

Description

This function is not recommended since a K-nearest neighbors approach with louvain clustering is a more generalizable approach to clustering. See the ??make_umap and ??knn_network_from_umap functions for more details. This function calculates the pairwise spearman correlaton for all anti-carbohydrate antibodies, then constructs a graph based on the spearman's rho and prunes weak edges based on p-value and spearmna's rho cutoffs.

Usage

make_correlation_network(
  iggdata,
  pcutoff = 0.05,
  rcutoff = 0.5,
  n_neighbor = 2
)

Arguments

iggdata	data frame of iggdata
pcutoff	p-value cutoff for edge removal from graph. Default = 0.05
rcutoff	Spearman's rho cutoff for edge removal from graph. Default = 0.5
n_neighbor	keep edges based on number of nearest neighbors. Default = 2

Value

igraph object

pmtran5884/Glycancc documentation built on Oct. 5, 2022, 10:14 a.m.