anota2seqRun: Wrapper for several functions which provide one-step analysis

In ChrOertlin/anota2seq: Generally applicable transcriptome-wide analysis of translational efficiency using anota2seq

Description

anota2seqRun is a wrapper function running the following steps of the anota2seq workflow: assessing model assumptions (by calling anota2seqPerformQC and anota2seqResidOutlierTest), performing analysis of changes in translational efficiency leading to altered protein levels or buffering; and differential expression of translated mRNA (e.g. polysome-associated mRNA or RPF) and total mRNA (by calling anota2seqAnalyze sequentially with analysis parameter set to "translation", "buffering", "translated mRNA", "total mRNA" respectively). Gene filtering is performed by calling anota2seqSelSigGenes and anota2seqRegModes categorizes regulated genes into regulatory modes (mRNA abundance or changes in translational efficiency leading to altered protein levels or buffering).

Usage

anota2seqRun(Anota2seqDataSet, contrasts = NULL, performQC = TRUE,
  onlyGroup = FALSE, performROT = TRUE, generateSingleGenePlots = FALSE,
  analyzeBuffering = TRUE, analyzemRNA = TRUE, thresholds = NULL,
  useRVM = TRUE, correctionMethod = "BH", useProgBar = TRUE)

Arguments

Anota2seqDataSet	Object of class Anota2seqDataSet
contrasts	If NULL (default), the contrasts will be created automatically and may not correspond to those of interest. It is therefore possible to use custom contrasts using this parameter. The input should be a matrix with row names corresponding to treatments (i.e. those in phenoVec or the treatment column of the SummarizedExperiment annotation) and columns corresponding to the different contrasts of interest (additional details on how to set up the contrast matrix is indicated in details section).
performQC	Boolean that defaults to TRUE. Used to specify if the anota2seqPerformQC function should be run.
onlyGroup	Boolean parameter of the anota2seqPerformQC function (default: FALSE). In anota2seqPerformQC, it is possible to suppress the omnibus interaction analysis and only perform the omnibus treatment analysis. Typically, when the data contains less than 3 samples in each sample class and more than 2 sample classes, the interaction analysis cannot be performed but the onlyGroup mode (i.e. with onlyGroup = TRUE) can be used to assess omnibus group effects.
performROT	Boolean that defaults to TRUE. Used to specify if the anota2seqResidOutlierTest function should be run.
generateSingleGenePlots	Should the single gene graphical outputs from the anota2seqPerformQC and anota2seqResidOulierTest functions be generated. Default is set to FALSE.
analyzeBuffering	Boolean that defaults to TRUE. Used to specify if changes in translational efficiency leading to buffering should be be analyzed.
analyzemRNA	Boolean that defaults to TRUE. Used to specify if translated mRNA (e.g. polysome-associated mRNA or RPFs) and total mRNA should be analyzed.
thresholds	A list containing thresholds that are applied during filtering of several parameters as described for the anota2seqSelSigGenes function. This list can contain the following name slots and if different from the below default values will update such defaults: minSlopeTranslation: The output can be filtered so that identifiers whose identified slopes in analysis of changes in translational efficiency leading to altered protein levels are too small can be excluded. Default is -1 i.e. excludes identifiers with a slope <(-1). maxSlopeTranslation: The output can be filtered so that identifiers whose identified slopes in analysis of changes in translational efficiency leading to altered protein levels are too large can be excluded. Default is 2 i.e. excludes identifiers with a slope > 2. minSlopeBuffering: The output can be filtered so that identifiers whose identified slopes in analysis of changes in translational efficiency leading to buffering are too small can be excluded. Default is -2 i.e. excludes identifiers with a slope <(-2). maxSlopeBuffering: The output can be filtered so that genes whose identified slopes in analysis of changes in translational efficiency leading to buffering are too large can be excluded. Default is 1 i.e. excludes identifiers with a slope > 1. maxPAdj: The output can be filtered based on adjusted p-values from the anota2seq analysis (i.e. smaller compared to assigned value). If useRVM is set to TRUE, filtering will be performed on RVM values, otherwise filtering will be performed on non-RVM values. Multiple testing adjustment method is set by argument correctionMethod. Default is 0.15. maxP: The output can be filtered based on unadjusted p-values from the anota2seq analysis (i.e. smaller compared to assigned value). If useRVM is set to TRUE, filtering will be performed on RVM values, otherwise filtering will be performed on non-RVM values. Default is NULL, i.e. no filtering. minEff: The output can be filtered based on minimum effect for inclusion. The value is applied both to negative and positive effects: e.g. a value of 1 will evaluate if the effects are >1 OR <(-1). Default is NULL i.e. no filtering based on effect. deltaPT: The output can be filtered based on the mean log2(translated mRNA data [e.g. polysome-associated mRNA or RPFs] / total mRNA data) difference between treatments. The treatments are defined by the selected contrast. Default is log2(1.2). Only applied when analysis is set to "translation". deltaTP: The output can be filtered based on the mean log2(total mRNA / translated mRNA data [e.g. polysome-associated mRNA or RPFs]) difference between treatments. The treatments are defined by the selected contrast. Default is log2(1.2). Only applied when analysis is set to "buffering". deltaP: The output can be filtered based on a minimum effect between the treatment and control groups in polysome-associated mRNA. Default is NULL (i.e. no filtering). Use when analysis parameter is set to "translation" (i.e. changes in translational efficiency leading to altered protein levels) or "translated mRNA" deltaT: The output can be filtered based on a minimum effect between the treatment and control groups in total mRNA. Default is NULL (i.e. no filtering). Use when analysis parameter is set to "buffering" or "total mRNA".
useRVM	Should the Random Variance Model be applied. Default is TRUE.
correctionMethod	Correction for multiple testing method. This parameter can be set to "Bonferroni", "Holm", "Hochberg", "SidakSS", "SidakSD", "BH", "BY", "ABH" or "TSBH" as implemented in the multtest package or "qvalue" as implemented in the qvalue package. Default is "BH".
useProgBar	Should the progress bar be shown. Default is TRUE, show progress bar.

Details

At the step of analysis, by default (i.e. with contrasts = NULL) the order of the sample classes which are used to calculate differences between treatments will be in alphabetical order. To change the directionality of the contrasts (e.g. treatment b vs treatment a instead of treatment a vs treatment b) or to generate a custom set of contrasts when more than 2 treatments are included, a contrast matrix can be supplied to the "contrasts" parameter described above. The row names should be specified as indicated above. The contrasts are coded by using e.g. -1 for group a, 0 for group b and 1 for group c to compare group a and c; -2 for group a, 1 for group b and 1 for group c to compare group a to b & c. Each column of the contrast matrix should sum to 0 and to analyze orthagonal contrasts the products of all pairwise rows should sum to 0. The results in the Anota2seqDataSet object will follow the order of the contrasts (i.e. results for e.g. contrast 1 will correspond to the contrasts specified in column 1 of the contrast matrix).

Value

An Anota2seqDataSet containing normalized data, model covariates and contrasts as well as outputs of all functions called by anota2seqRun. anota2seqRun will also output all diagnostic plots provided by anota2seqPerformQC, anota2seqResidOutlierTest and anota2seqAnalyze.

See Also

anota2seqPerformQC, anota2seqResidOutlierTest, anota2seqAnalyze, anota2seqSelSigGenes anota2seqRegModes

Examples

data(anota2seq_data)
# Initialize Anota2seqDataSet
Anota2seqDataSet <- anota2seqDataSetFromMatrix(
    dataP = anota2seq_data_P[1:100,],
    dataT = anota2seq_data_T[1:100,],
    phenoVec = anota2seq_pheno_vec,
    dataType = "RNAseq",
    normalize = TRUE)
# Perform anota2seqRun function
# The quality control and residual outlier testing are not 
# performed in order to limit the running time of this example, but the model 
# assumptions should be assessed (see help of anota2seqPerformQC)
Anota2seqDataSet <- anota2seqRun(Anota2seqDataSet,
    performQC = FALSE, 
    performROT = FALSE, 
    useProgBar = FALSE)
    
## Not run:     
# Example to build a custom contrast matrix
# For the purpose of this example, we will use the first 6 samples of the 
# simulated data provided with the package together with the following "dummy"
# sample classes:
phenoVec <- c("a","a","b","b","c","c")
contrastsEx_ads <- anota2seqDataSetFromMatrix(
    dataP = anota2seq_data_P[1:300, 1:6],
    dataT = anota2seq_data_T[1:300, 1:6],
    phenoVec = phenoVec,
    dataType = "RNAseq",
    normalize = TRUE)

# Get the levels of the phenoVec, these will be ordered as in anota2seq
phenoLev <- levels(as.factor(phenoVec))
# Construct the matrix with appropriate nrow and ncol
myContrast <- matrix(nrow =length(phenoLev),ncol=length(phenoLev)-1)
# Set the phenoLev as rownames for your contrast matrix
rownames(myContrast) <- phenoLev
# Now indicate the contrasts you want to analyse as explained above
# Compare a to c
myContrast[,1] <- c(-1,0,1)
# Compare a to b& c
myContrast[,2] <- c(2,-1,-1)
myContrast
#   [,1] [,2]
# a   -1    2
# b    0   -1
# c    1   -1
# The custom contrast matrix can then be used as input of anota2seqRun. Because 
# these data have only 2 samples per sample class, the onlyGroup mode of 
# anota2seqPerformQC is the only available mode for assessment of model 
# assumptions so we also set onlyGroup to TRUE.
contrastsEx_ads <- anota2seqRun(contrastsEx_ads,
                                contrasts = myContrast, 
                                performQC = FALSE, 
                                performROT = FALSE,
                                onlyGroup = TRUE,
                                thresholds = list(
                                    maxPAdj = 0.25,
                                    deltaPT = log2(2))) 
                                    
## End(Not run)

ChrOertlin/anota2seq documentation built on Aug. 4, 2021, 2:17 p.m.