amp_network: Correlation Network

In julieklessner/cornetwork: Visualisation of correlations

Description

This function allows you to build a network based on correlations

Usage

amp_network(data, threshold.cor = 0.8, threshold.abund = 0.01,
  threshold.obs = 0.1, threshold.pval = 0.05, show.label = FALSE,
  scale.abund = FALSE, node.size = 5, scale.cor = FALSE,
  edge.size = 0.5, highlight.OTU = NULL, highlight.color = NULL,
  highlight.size = 10, highlight.label = NULL, tax.display = "Phylum",
  tax.add = NULL, tax.class = NULL, tax.empty = "best",
  add.tax.info = FALSE, show.edge.label = FALSE, correlation = "Pearson",
  return.output = "top", show.top = 25)

Arguments

data	(required) A phyloseq object.
threshold.cor	Threshold value for the correlation (default: 0.8).
threshold.abund	Threshold value for relative abundance (default: 0.01).
threshold.obs	Threshold value for the average relative abundance per observation (default: 0.1).
threshold.pval	Threshold value for the p-value (default: 0.05).
show.label	Show vertex names as labels (default: FALSE).
scale.abund	Scale the node size according to average abundance (default: FALSE).
node.size	Size of nodes if scale.abund=FALSE (default: 5)
scale.cor	Scale the edge color according to strength of correlation (default: FALSE).
edge.size	Size of egdes if scale.cor=FALSE (default: 0.5)
highlight.OTU	Highlight OTUs using see-throgh nodes, e.g. c("MiDAS_37", "MiDAS_73") (default: none).
highlight.color	Color of highlights, e.g. c("purple", "green") (default: none).
highlight.size	Size of highlights (default: 10).
highlight.label	Highlight OTUs using node labels, e.g. c("MiDAS_37", "MiDAS_73") (default:none).
tax.display	The taxonomic level that the data should be displayed (default: "Phylum").
tax.class	Converts a specific phyla to class level instead, e.g. "p__Proteobacteria" (default: "none").
tax.empty	Either "remove" OTUs without taxonomic information, add "best" classification or add the "OTU" name (default: "best").
add.tax.info	Show taxonomic information (default: FALSE).
show.edge.label	Show correlation values as edge labels (default: FALSE).
correlation	Choose the method for calculation of correlation from Pearson, Spearman, SparCC, CClasso or Rebacca (default: "Pearson").
return.output	Either "top" or "total" or "plot" (default: "top").
show.top	Number of correlations returned if return.output is set to top (default: 25).

julieklessner/cornetwork documentation built on May 20, 2019, 4:22 a.m.