corMatrix: Correlation Matrix and tree

In AurelieGuilbault/VIQCing: Data Processing For Metabolomic Data

Description

Creates the correlation matrix and tree for a metabolomic dataset. Produces the following files:

• <filename>.pdf, containing the correlation matrix

• <filename>_pairs.txt, containing the correlation pairs above <upperLimit> or less than <lowerLimit>;

Usage

corMatrix(filename, output = paste0(strsplit(filename, ".", fixed =
  TRUE)[[1]][1], "_Correlation.pdf"), na = F, landscape = F,
  dendrogram = T, heatmap = T, compound = NULL, metabolite = NULL,
  sampleStart = 3, corDigit = 3, testType = "spearman",
  plotWidth = 0.5, plotHeigth = 0.2, textSize = 0.3,
  upperLimit = 0.6, lowerLimit = -0.6)

Arguments

filename	name of the datafile Must contain: a column "Compound"; a column "Metabolite"; and all the columns sample from <sampleStart> to the end of the file; the rest doesn't matter and the names are optional, as long as the column position is entered.;
output	default <dataset filename>.pdf, name of the pdf file;
na	default FALSE, FALSE removes the untargeted meatabolite;
landscape	default FALSE, orientation of the pdf file;
dendrogram	default TRUE, presence of the correlation dendrogram;
heatmap	default TRUE, presence of the correlation matrix heatmap;
compound	default NULL, position of the compound column if named otherwise;
metabolite	default NULL, position of the metabolite column if named otherwise;
sampleStart	default 3, 1st column of the actual data.
corDigit	default 3,number of digit for the correlation coefficient;
testType	default spearman, type of correlation test. Can be "spearman"or "pearson";
plotWidth	default 0.5, width of the top space above the heatmap;
plotHeigth	default 0.2, width of the side space beside the heatmap;
textSize	default 2, size of the text in the heatmap tiles and the dendrogram;
upperLimit	default 0.6, positive correlation limit to output the metabolite pair;
lowerLimit	default -0.6, negative correlation limit to output the metabolite pair.

Details

The significance of the correlation is noted with a "*" in the tile. Conditions for the Pearson's Correlation Test:

• Independent samples;

• normal distribution of the data; *

*Since this condition is not true for most metabolite, the test is set as a Spearman's rank Correlation test by default.*

Value

the rcorr objet, with:

• r The correlation matrix

• P The P-value matrix

See Also

• reshape2 packagehttps://www.rdocumentation.org/packages/reshape2

• Hmisc package https://www.rdocumentation.org/packages/Hmisc

• gplot package https://www.rdocumentation.org/packages/gplot

Examples

for a dataset with the following header ; Compound, m/z, Metabolite, RT, Sample #1, ...
corMatrix("dummySet.tsv", sampleStart=5, testType="spearman")

for a dataset with the following header ; compound, m/z, metabolite, RT, Sample #1, ...
corMatrix("dummySet.tsv", compound=1, metabolite=3, sampleStart=5, testType="spearman")

AurelieGuilbault/VIQCing documentation built on Oct. 30, 2019, 5:02 a.m.