choose_gaussians: Fit a Gaussian mixture model to a co-elution profile

In PrInCE: Predicting Interactomes from Co-Elution

Description

Fit mixtures of one or more Gaussians to the curve formed by a chromatogram profile, and choose the best fitting model using an information criterion of choice.

Usage

choose_gaussians(chromatogram, points = NULL, max_gaussians = 5,
  criterion = c("AICc", "AIC", "BIC"), max_iterations = 10,
  min_R_squared = 0.5, method = c("guess", "random"),
  filter_gaussians_center = TRUE, filter_gaussians_height = 0.15,
  filter_gaussians_variance_min = 0.1,
  filter_gaussians_variance_max = 50)

Arguments

chromatogram	a numeric vector corresponding to the chromatogram trace
points	optional, the number of non-NA points in the raw data
max_gaussians	the maximum number of Gaussians to fit; defaults to 5. Note that Gaussian mixtures with more parameters than observed (i.e., non-zero or NA) points will not be fit.
criterion	the criterion to use for model selection; one of "AICc" (corrected AIC, and default), "AIC", or "BIC"
max_iterations	the number of times to try fitting the curve with different initial conditions; defaults to 10
min_R_squared	the minimum R-squared value to accept when fitting the curve with different initial conditions; defaults to 0.5
method	the method used to select the initial conditions for nonlinear least squares optimization (one of "guess" or "random"); see make_initial_conditions for details
filter_gaussians_center	true or false: filter Gaussians whose centres fall outside the bounds of the chromatogram
filter_gaussians_height	Gaussians whose heights are below this fraction of the chromatogram height will be filtered. Setting this value to zero disables height-based filtering of fit Gaussians
filter_gaussians_variance_min	Gaussians whose variance is below this threshold will be filtered. Setting this value to zero disables filtering.
filter_gaussians_variance_max	Gaussians whose variance is above this threshold will be filtered. Setting this value to zero disables filtering.

Value

a list with five entries: the number of Gaussians used to fit the curve; the R^2 of the fit; the number of iterations used to fit the curve with different initial conditions; the coefficients of the fit model; and the curve predicted by the fit model.

Examples

data(scott)
chrom <- clean_profile(scott[1, ])
gauss <- choose_gaussians(chrom, max_gaussians = 3)

PrInCE documentation built on Nov. 8, 2020, 6:34 p.m.