eig_norm2: EigenMS normalization

In ProteoMM: Multi-Dataset Model-based Differential Expression Proteomics Analysis Platform

Description

Eliminate the effects of systematic bias identified in eig_norm1() Ref: "Normalization of peak intensities in bottom-up MS-based proteomics using singular value decomposition" Karpievitch YV, Taverner T et al. 2009, Bioinformatics Ref: "Metabolomics data normalization with EigenMS" Karpievitch YK, Nikolic SB, Wilson R, Sharman JE, Edwards LM Submitted to PLoS ONE.

Usage

eig_norm2(rv)

Arguments

rv	return value from the eig_norm1 if user wants to change the number of bias trends that will be eliminated h.c in rv should be updates to the desired number

Value

A structure with multiple components

normalized

matrix of normalized abundances with 2 columns of protein and peptdie names

norm_m

matrix of normalized abundances, no extra columns

eigentrends

trends found in raw data, bias trends up to h.c

norm.svd

trends in normalized data, if one wanted to plot at later time

exPeps

peptides excluded due to not enough peptides or exception in fitting a linear model

Examples

data(mm_peptides)
head(mm_peptides)
# different from parameter names as R uses outer name
# spaces if variable is undefined
intsCols = 8:13
metaCols = 1:7 # reusing this variable
m_logInts = make_intencities(mm_peptides, intsCols)
m_prot.info = make_meta(mm_peptides, metaCols)
m_logInts = convert_log2(m_logInts)
grps = as.factor(c('CG','CG','CG', 'mCG','mCG','mCG'))

set.seed(123) # set for repoducubility of eig_norm1
mm_m_ints_eig1 = eig_norm1(m=m_logInts,treatment=grps,prot.info=m_prot.info)
mm_m_ints_eig1$h.c # check the number of bias trends detected
mm_m_ints_norm = eig_norm2(rv=mm_m_ints_eig1)

ProteoMM documentation built on Nov. 8, 2020, 5:57 p.m.