mixtureHeatMap: Heat map of mixture CPA errors

In mooredf22/protlocassign: Estimation of protein distribution from quantitative subcellular fractionation experiments using a constrained proportional assignment model

Heat map of mixture CPA errors

Description

Produce heat map of errors in CPA estimates for fitting simulated proteins distributed between all pairs of compartments using different types of profiles (RSA, NSA, and Acup; linear and log2-transformed). Formatted for use with up to eight compartments , which produces 28 (8 chose 2) pairs of combinations of compartments. Also writes a table of total assignment errors for each type of transformation.

Usage

mixtureHeatMap(
  Acup,
  totProt,
  NstartMaterialFractions = 6,
  eps = 0.001,
  errorListIn = NULL,
  errorListReturn = FALSE
)

Arguments

Acup	Acup profiles of reference compartments
totProt	vector of total protein amounts (derived from a given amount of starting material) in each of the fractions comprising the profile
NstartMaterialFractions	Number of fractions that reconstitute the starting material, e.g., a complete set of differential centrifugation fractions. For experiment AT5, it is 6 ( N, M, L1, L2, P, and S).
eps	small value to add so that log argument is greater than zero
errorListIn	If not NULL, pre-computed errors are passed to the function to save computing time for display purposes. The format must be as in the $errorList component in the output. Default is NULL
errorListReturn	If TRUE, return data frame of mixture errors if FALSE (the default) return matrix of total errors

Value

Plot of heat map of errors of CPA estimates for pairs of simulated protein mixtures represented by different types of profiles and a table of total errors for each type of transformation. Errors are displayed in a 3x2 output where the top row of the left, middle and right columns represent RSA, NSA, and Acup profiles, respectively and the bottom row represents a log2-transformation of these profiles. By default return a matrix total errors. If errorListShow=TRUE, return data frame of mixture errors

Examples

data(totProtAT5)
data(refLocProfAcup)
par(mfrow=c(1,1))
errorResult <- mixtureHeatMap(Acup=refLocProfAcup, totProt=totProtAT5)
round(head(errorResult), digits=3)

# Note that the profile of one protein, AIF1, contains missing values
# which causes the cpa routine to generate a nonconvergence message 

mooredf22/protlocassign documentation built on Sept. 13, 2023, 3:57 p.m.