filt.hierClust: filt.hierClust

filt.hierClustR Documentation

filt.hierClust

Description

This function takes as input a square similarity matrix and searches for clusters of samples with strong associations and extracts the sub matrix with the closely related sampless. Only positive correlations are considered here.

Usage

filt.hierClust(
  mat.rho,
  hclust.method = "ward",
  margins = c(6, 6),
  side.col.c = NULL,
  side.col.r = NULL,
  size = 10,
  plot = TRUE,
  filt = 0.5
)

Arguments

mat.rho

: square correlation matrix with ids (can be used for also other than just samples)

hclust.method

: the hierarchical clustering method, by default it is the ward.D method

margins

: change margins of the graph (default = c(6, 6))

side.col.c

: a vector of colors to be applied in the columns, usually depincting a class

side.col.r

: a vector of colors to be applied in the rows, usually depincting a class

size

: the number of samples in the resulting ordered matrix

plot

: logical default TRUE. It will plot the heatmap of the similarity with the hierchical clustering

filt

: default is 0.5 and is the filtering threshold to be applied

Details

filt.hierClust

Value

it will return a matrix with samples in rows and their closely related ones on the columns along with the correlation score.

Author(s)

Emmanuelle Le Chatelier & Edi Prifti


eprifti/momr documentation built on Sept. 27, 2022, 3:36 a.m.