plotload: Plotting PCA or OPLS loadings

In kimsche/MetaboMate: All you need for NMR-based metabolic profiling with R!

Description

Plotting PCA or OPLS loadings

Usage

plotload(model, X, ppm, shift = c(0, 10), pc = 1,
  type = c("Statistical reconstruction", "Backscaled"), title = "")

Arguments

model	PCA or OPLS model generated via MetaboMate package functions.
X	Input matrix with rows representing spectra
ppm	ppm variable
shift	ppm region to visualise.
pc	index of principal component to visualise, set to 1 if input model is OPLS
type	Type of loadings visualisation, either 'Statistical reconstruction' or 'Backscaled' (see Details).
title	Plot title.

Details

OPLS: If type='Statistical reconstruction' the function calculates the covariance (y axis) and Pearson's correlation (colouring) of the predictive OPLS scores with each X variable (x axis is ppm variable). If type='Backscaled' the OPLS loadings are backscaled with X feature standard deviations. Results are plotted over ppm, coloured according to OPLS model weights. Often, the latter method visualises model importance more robust due to the presence of false positive correlations. PCA: Function always calculates the statistical recostruction.

Author(s)

Torben Kimhofer [e-mail]

References

Cloarec, O., et al. (2005). Evaluation of the Orthogonal Projection on Latent Structure Model Limitations Caused by Chemical Shift Variability and Improved Visualization of Biomarker Changes in 1H NMR Spectroscopic Metabonomic Studies. Analytical Chemistry 77.2, 517-26.

See Also

pca opls PCA_MetaboMate-class OPLS_MetaboMate-class

kimsche/MetaboMate documentation built on Aug. 8, 2020, 1:14 a.m.