Package 'erccdashboard'

Title: Assess Differential Gene Expression Experiments with ERCC Controls
Description: Technical performance metrics for differential gene expression experiments using External RNA Controls Consortium (ERCC) spike-in ratio mixtures.
Authors: Sarah Munro, Steve Lund
Maintainer: Sarah Munro <[email protected]>
License: GPL (>=2)
Version: 1.41.0
Built: 2024-10-31 06:07:19 UTC
Source: https://github.com/bioc/erccdashboard

Help Index

Annotate signal-abundance and ratio-abundance plots with LODR

Description

Annotate signal-abundance and ratio-abundance plots with LODR

Usage

annotLODR(exDat)

Arguments

exDat

list, contains input data and stores analysis results

Examples

data(SEQC.Example)
 
exDat <- initDat(datType="array", isNorm=FALSE, 
                 exTable=UHRR.HBRR.arrayDat,
                 filenameRoot="testRun", sample1Name="UHRR",
                 sample2Name="HBRR", erccmix="RatioPair", 
                 erccdilution = 1, spikeVol = 50, 
                 totalRNAmass = 2.5*10^(3), choseFDR=0.01)
                 
exDat <- est_r_m(exDat)
                  
exDat <- dynRangePlot(exDat)

exDat <- geneExprTest(exDat)

exDat <- estLODR(exDat, kind="ERCC", prob=0.9)

exDat <- annotLODR(exDat)

exDat$Figures$maPlot

Produce signal-abundance plot to evaluate dynamic range

Description

Produce signal-abundance plot to evaluate dynamic range

Usage

dynRangePlot(exDat, allPoints, labelReps)

Arguments

exDat

list, contains input data and stores analysis results
allPoints

boolean, default is false, means of replicates will be plotted. If true then all replicates will be plotted as individual points.
labelReps

boolean, default is false. If true then replicates will be labeled.

Examples

data(SEQC.Example)

exDat <- initDat(datType="count", isNorm=FALSE, exTable=MET.CTL.countDat, 
                 filenameRoot="testRun", sample1Name="MET", 
                 sample2Name="CTL", erccmix="RatioPair", 
                 erccdilution=1/100, spikeVol=1, totalRNAmass=0.500,
                 choseFDR=0.1)
                 
exDat <- est_r_m(exDat)
                  
exDat <- dynRangePlot(exDat, allPoints="FALSE", labelReps ="FALSE")

exDat$Figures$dynRangePlot

ERCC data

Description

Contains 2 data frames: ERCCDef and ERCCMix1and2

Usage

data(ERCC)

Examples

data(ERCC)

ERCCDef dataframe

Description

ERCC transcript lengths and GC content

Format

A data frame with 96 observations on the following 3 variables.

Feature

a factor vector

Length

a numeric vector

GC

a numeric vector

Details

Length and GC content of all 96 ERCC controls in NIST SRM 2374

Source

http://tinyurl.com/erccsrm

ERCCMix1and2 dataframe

Description

Ambion RatioPair ERCC Mixtures

Format

A data frame with 96 observations on the following 4 variables.

ERCC.AMB.Expected

a factor vector of all 96 ERCC control IDs

Subpool

a factor vector of the ERCC Ratios in each Subpool with levels 4:1 1:1 1:1.5 1:2

Mix1Conc.Attomoles_ul

a numeric vector of the ERCC concentrations in Mix 1

Mix2Conc.Attomoles_ul

a numeric vector of the ERCC concentrations in Mix 2

Source

http://www.lifetechnologies.com/order/catalog/product/4456739

Produce Receiver Operator Characteristic (ROC) Curves and AUC statistics

Description

Produce Receiver Operator Characteristic (ROC) Curves and AUC statistics

Usage

erccROC(exDat)

Arguments

exDat

list, contains input data and stores analysis results

Examples

data(SEQC.Example)

exDat <- initDat(datType="array", isNorm=FALSE, 
                 exTable=UHRR.HBRR.arrayDat,
                 filenameRoot="testRun", sample1Name="UHRR",
                 sample2Name="HBRR", erccmix="RatioPair", 
                 erccdilution = 1, spikeVol = 50, 
                 totalRNAmass = 2.5*10^(3), choseFDR=0.01)
                 
exDat <- est_r_m(exDat)
                  
exDat <- dynRangePlot(exDat)

exDat <- geneExprTest(exDat)

exDat <- erccROC(exDat)

exDat$Figures$rocPlot

Estimate the mRNA fraction differences for the pair of samples using replicate data

Description

Estimate the mRNA fraction differences for the pair of samples using replicate data

Usage

est_r_m(exDat)

Arguments

exDat

list, contains input data and stores analysis results

Details

This is the first function to run after an exDat structure is initialized using initDat, because it is needed for all additional analysis. An r_m of 1 indicates that the two sample types under comparison have similar mRNA fractions of total RNA. The r_m estimate is used to adjusted the expected ERCC mixture ratios in this analysis and may indicate a need for a different sample normalization approach.

Examples

data(SEQC.Example)

exDat <- initDat(datType="count", isNorm = FALSE, exTable=MET.CTL.countDat, 
                 filenameRoot = "testRun",sample1Name = "MET",
                 sample2Name = "CTL", erccmix = "RatioPair", 
                 erccdilution = 1/100, spikeVol = 1, totalRNAmass = 0.500,
                 choseFDR = 0.1)

exDat <- est_r_m(exDat)

Estimate Limit of Detection of Ratios (LODR)

Description

Estimate Limit of Detection of Ratios (LODR)

Usage

estLODR(exDat, kind = "ERCC", prob = 0.9)

Arguments

exDat

list, contains input data and stores analysis results
kind

"ERCC" or "Sim"
prob

probability, ranging from 0 - 1, default is 0.9

Details

This is the function to estimate a limit of detection of ratios (LODR) for a a chosen probability and threshold p-value for the fold changes in the ERCC control ratio mixtures.

Examples

data(SEQC.Example)
 
exDat <- initDat(datType="array", isNorm=FALSE, 
                 exTable=UHRR.HBRR.arrayDat,
                 filenameRoot="testRun", sample1Name="UHRR",
                 sample2Name="HBRR", erccmix="RatioPair", 
                 erccdilution = 1, spikeVol = 50, 
                 totalRNAmass = 2.5*10^(3), choseFDR=0.01)
                 
exDat <- est_r_m(exDat)
                  
exDat <- dynRangePlot(exDat)

exDat <- geneExprTest(exDat)

exDat <- estLODR(exDat, kind = "ERCC", prob = 0.9)

exDat$Figures$lodrERCCPlot

Prepare differential expression testing results for spike-in analysis

Description

Prepare differential expression testing results for spike-in analysis

Usage

geneExprTest(exDat)

Arguments

exDat

list, contains input data and stores analysis results

Details

This function wraps the edgeR differential expression testing package for datType = "count" or uses the limma package for differential expression testing if datType = "array". Alternatively, for count data only, if correctly formatted DE test results are provided, then geneExprTest will bypass DE testing (with reduced runtime).

Examples

data(SEQC.Example)

exDat <- initDat(datType="array", isNorm=FALSE, 
                 exTable=UHRR.HBRR.arrayDat,
                 filenameRoot="testRun", sample1Name="UHRR",
                 sample2Name="HBRR", erccmix="RatioPair", 
                 erccdilution = 1, spikeVol = 50, 
                 totalRNAmass = 2.5*10^(3), choseFDR=0.01)
                 
exDat <- est_r_m(exDat)
                  
exDat <- dynRangePlot(exDat)

exDat <- geneExprTest(exDat)

Initialize the exDat list

Description

Initialize the exDat list

Usage

initDat(
  datType = NULL,
  isNorm = FALSE,
  exTable = NULL,
  repNormFactor = NULL,
  filenameRoot = NULL,
  sample1Name = NULL,
  sample2Name = NULL,
  erccmix = "RatioPair",
  erccdilution = 1,
  spikeVol = 1,
  totalRNAmass = 1,
  choseFDR = 0.05,
  ratioLim = c(-4, 4),
  signalLim = c(-14, 14),
  userMixFile = NULL
)

Arguments

datType

type is "count" or "array", unnormalized data is expected (normalized data may be accepted in future version of the package). Default is "count" (integer count data),"array" is unnormalized fluorescent intensities from microarray fluorescent intensities (not log transformed or normalized)
isNorm

default is FALSE, if FALSE then the unnormalized input data will be normalized in erccdashboard analysis. If TRUE then it is expected that the data is already normalized
exTable

data frame, the first column contains names of genes or transcripts (Feature) and the remaining columns are counts for sample replicates spiked with ERCC controls
repNormFactor

optional vector of normalization factors for each replicate, default value is NULL and 75th percentile normalization will be applied to replicates
filenameRoot

string root name for output files
sample1Name

string name for sample 1 in the gene expression experiment
sample2Name

string name for sample 2 in the gene expression experiment
erccmix

Name of ERCC mixture design, "RatioPair" is default, the other option is "Single"
erccdilution

unitless dilution factor used in dilution of the Ambion ERCC spike-in mixture solutions
spikeVol

volume in microliters of diluted ERCC mix spiked into the total RNA samples
totalRNAmass

mass in micrograms of total RNA spiked with diluted ERCC mixtures
choseFDR

False Discovery Rate for differential expression testing , default is 0.05
ratioLim

Limits for ratio axis on MA plot, default is c(-4,4)
signalLim

Limits for signal axis on dynamic range plot, default is c(-14,14)
userMixFile

optional filename input, default is NULL, if ERCC control ratio mixtures other than the Ambion product were used then a userMixFile can be used for the analysis

Examples

data(SEQC.Example)

exDat <- initDat(datType="count", isNorm = FALSE, exTable=MET.CTL.countDat, 
                 filenameRoot = "testRun",sample1Name = "MET",
                 sample2Name = "CTL", erccmix = "RatioPair", 
                 erccdilution = 1/100, spikeVol = 1, totalRNAmass = 0.500,
                 choseFDR = 0.1)
summary(exDat)

Generate MA plots with or without annotation using LODR estimates

Description

Generate MA plots with or without annotation using LODR estimates

Usage

maSignal(exDat, alphaPoint = 0.8, r_mAdjust = TRUE, replicate = TRUE)

Arguments

exDat

list, contains input data and stores analysis results
alphaPoint

numeric value, for alpha (transparency) for plotted points, range is 0 - 1
r_mAdjust

default is TRUE, if FALSE then the r_m estimate will not used to offset dashed lines for empirical ratios on figure
replicate

default is TRUE, if FALSE then error bars will not be produced

Examples

data(SEQC.Example)

exDat <- initDat(datType="array", isNorm=FALSE, 
                 exTable=UHRR.HBRR.arrayDat,
                 filenameRoot="testRun", sample1Name="UHRR",
                 sample2Name="HBRR", erccmix="RatioPair", 
                 erccdilution = 1, spikeVol = 50, 
                 totalRNAmass = 2.5*10^(3), choseFDR=0.01)
                 
exDat <- est_r_m(exDat)
                  
exDat <- dynRangePlot(exDat)

exDat <- geneExprTest(exDat)
# generate MA plot without LODR annotation
exDat <- maSignal(exDat)

exDat$Figures$maPlot

exDat <- estLODR(exDat, kind = "ERCC", prob = 0.9)

# Include LODR annotation
exDat <- annotLODR(exDat)

exDat$Figures$maPlot

Rat toxicogenomics count data

Description

RNA-Seq count data from Methimazole and Control rat biological replicates

Format

A data frame with 16590 observations of the following 7 variables.

Feature

a factor vector of all Endogenous and ERCC transcripts in the experiment

MET_1

a numeric vector of counts from Methimazole treatment biological replicate 1

MET_2

a numeric vector of counts from Methimazole treatment biological replicate 2

MET_3

a numeric vector of counts from Methimazole treatment biological replicate 3

CTL_1

a numeric vector of counts from Control biological replicate 1

CTL_2

a numeric vector of counts from Control biological replicate 2

CTL_3

a numeric vector of counts from Control biological replicate 3

Rat toxicogenomics total read data

Description

Total reads per biological replicate from FASTQ files

Format

The format is: int [1:6] 41423502 46016148 44320280 38400362 47511484 33910098

Run default erccdashboard analysis of ERCC control ratio mixtures

Description

Run default erccdashboard analysis of ERCC control ratio mixtures

Usage

runDashboard(
  datType = NULL,
  isNorm = FALSE,
  exTable = NULL,
  repNormFactor = NULL,
  filenameRoot = NULL,
  sample1Name = NULL,
  sample2Name = NULL,
  erccmix = "RatioPair",
  erccdilution = 1,
  spikeVol = 1,
  totalRNAmass = 1,
  choseFDR = 0.05,
  ratioLim = c(-4, 4),
  signalLim = c(-14, 14),
  userMixFile = NULL
)

Arguments

datType

type is "count" (RNA-Seq) or "array" (microarray), "count" is unnormalized integer count data (normalized RNA-Seq data will be accepted in an updated version of the package), "array" can be normalized or unnormalized fluorescent intensities from a microarray experiment.
isNorm

default is FALSE, if FALSE then the unnormalized input data will be normalized in erccdashboard analysis. If TRUE then it is expected that the data is already normalized
exTable

data frame, the first column contains names of genes or transcripts (Feature) and the remaining columns are expression measures for sample replicates spiked with ERCC controls
repNormFactor

optional vector of normalization factors for each replicate, default value is NULL and 75th percentile normalization will be applied to replicates
filenameRoot

string root name for output files
sample1Name

string name for sample 1 in the gene expression experiment
sample2Name

string name for sample 2 in the gene expression experiment
erccmix

Name of ERCC mixture design, "RatioPair" is default, the other option is "Single"
erccdilution

unitless dilution factor used in dilution of the Ambion ERCC spike-in mixture solutions
spikeVol

volume in microliters of diluted ERCC mix spiked into the total RNA samples
totalRNAmass

mass in micrograms of total RNA spiked with diluted ERCC mixtures
choseFDR

False Discovery Rate for differential expression testing
ratioLim

Limits for ratio axis on MA plot, default is c(-4,4)
signalLim

Limits for ratio axis on MA plot, default is c(-14,14)
userMixFile

optional filename input, default is NULL, if ERCC control ratio mixtures other than the Ambion product were used then a userMixFile can be used for the analysis

Examples

## Not run: 
data(SEQC.Example)
     
exDat <- runDashboard(datType = "count",isNorm = FALSE,
                     exTable = MET.CTL.countDat, 
                     filenameRoot = "COH.ILM",
                     sample1Name = "MET", sample2Name = "CTL", 
                     erccmix = "RatioPair", erccdilution = 1/100, 
                     spikeVol = 1, totalRNAmass = 0.500,choseFDR = 0.1)
                 
summary(exDat)

## End(Not run)

Save erccdashboard plots to a pdf file

Description

The function savePlots will save selected figures to a pdf file. The default is the 4 manuscript figures to a single page (plotsPerPg = "manuscript"). If plotsPerPg = "single" then each plot is placed on an individual page. If plotlist is not defined (plotlist = NULL) or if plotlist = exDat$Figures then all plots in exDat$Figures are printed to a PDF file.

Usage

saveERCCPlots(
  exDat,
  plotsPerPg = "main",
  saveas = "pdf",
  outName = NULL,
  plotlist = NULL,
  res = 200
)

Arguments

exDat

list, contains input data and stores analysis results
plotsPerPg

string, if "main" then the 4 main plots are printed to one page, if "single" then a single plot is printed per page from the plotlist argument
saveas

Choose file format from "pdf", "jpeg" or "png"
outName

Choose output file name, default will be fileName from exDat
plotlist

list, contains plots to print
res

Choose the file resolution

Examples

## Not run: 
data(SEQC.Example)
 
exDat <- initDat(datType="count", isNorm=FALSE, exTable=MET.CTL.countDat, 
                 filenameRoot="testRun", sample1Name="MET",
                 sample2Name="CTL", erccmix="RatioPair", 
                 erccdilution=1/100, spikeVol=1, totalRNAmass=0.500,
                 choseFDR=0.1)
                 
exDat <- est_r_m(exDat)
                  
exDat <- dynRangePlot(exDat)

exDat <- geneExprTest(exDat)

exDat <- erccROC(exDat)

exDat <- estLODR(exDat, kind="ERCC", prob=0.9)

exDat <- annotLODR(exDat)

#to print 4 main plots to a single page pdf file
saveERCCPlots(exDat, plotsPerPg = "main",saveas = "pdf")

#to print 4 plots to a jpeg file
saveERCCPlots(exDat, plotsPerPg = "main",saveas = "jpeg")

# or to create a multiple page pdf of all plots produced
saveERCCPlots(exDat, plotsPerPg = "single", plotlist = exDat$Figures)



## End(Not run)

Example data from SEQC project for erccdashboard analysis

Description

Contains the following 5 itemsL MET.CTL.countDat - Rat toxicogenomics count data MET.CTL.totalReads - Rat toxicogenomics total read data UHRR.HBRR.arrayDat - UHRR and HBRR Illumina BeadArray data UHRR.HBRR.countDat - UHRR and HBRR RNA-Seq Illumina count data UHRR.HBRR.totalReads - UHRR and HBRR sample total read data

Usage

data(SEQC.Example)

Examples

data(SEQC.Example)

UHRR and HBRR Illumina BeadArray data

Description

Unnormalized microarray data from Lab 13 of reference sample interlaboratory study

Format

A data frame with 17627 observations of the following 7 variables.

Feature

a factor vector of all Endogenous and ERCC transcripts in the experiment

UHRR_3

a numeric vector of fluorescence intensities from UHRR microarray technical replicate 1

UHRR_2

a numeric vector of fluorescence intensities from UHRR microarray technical replicate 2

UHRR_1

a numeric vector of fluorescence intensities from UHRR microarray technical replicate 3

HBRR_3

a numeric vector of fluorescence intensities from HBRR microarray technical replicate 1

HBRR_2

a numeric vector of fluorescence intensities from HBRR microarray technical replicate 2

HBRR_1

a numeric vector of fluorescence intensities from HBRR microarray technical replicate 3

UHRR and HBRR RNA-Seq Illumina count data

Description

RNA-Seq count data from UHRR and HBRR interlaboratory study library replicates

Format

A data frame with 43919 observations of the following 9 variables.

Feature

a character vector of all Endogenous and ERCC transcripts in the experiment

UHRR_1

a numeric vector of counts from UHRR library preparation replicate 1

UHRR_2

a numeric vector of counts from UHRR library preparation replicate 2

UHRR_3

a numeric vector of counts from UHRR library preparation replicate 3

UHRR_4

a numeric vector of counts from UHRR library preparation replicate 4

HBRR_1

a numeric vector of counts from HBRR library preparation replicate 1

HBRR_2

a numeric vector of counts from HBRR library preparation replicate 2

HBRR_3

a numeric vector of counts from HBRR library preparation replicate 3

HBRR_4

a numeric vector of counts from HBRR library preparation replicate 4

UHRR and HBRR sample total read data

Description

Total reads per library replicate from FASTQ files

Format

The format is: int [1:8] 138786892 256006510 199468322 431933806 247985592 219383270 251265814 257508210