In ipb-halle/MetFamily: MetFamily: Discovering Regulated Metabolite Families in Untargeted Metabolomics Studies

library(MetFamily)

Introduction

Some text about the scenario, data and MTBLS297.

Cool paper [@Treutler16DiscoveringRegulatedMetabolite].

Cool R package: r Githubpkg("ipb-halle/MetFamily")

Loading the data

First, we load the data and summarise it

# Load separate files

filePeakMatrixPath <- system.file("extdata/showcase/Metabolite_profile_showcase.txt", package = "MetFamily")

fileSpectra <- system.file("extdata/showcase/MSMS_library_showcase.msp", package = "MetFamily")

fileAnnotation <- system.file("extdata/testdata/canopus/canopusShort.txt", package = "MetFamily")

parameterSetPath <- system.file("extdata/testdata/parameterSet.RData",  package = "MetFamily" )
load(parameterSetPath)

resultObj <- convertToProjectFile(filePeakMatrixPath, 
                                 fileSpectra,
                                 parameterSet = parameterSet, 
                                 progress = FALSE)

lines <- sparseMatrixToString(matrixRows = resultObj$matrixRows, matrixCols = resultObj$matrixCols,
                              matrixVals = resultObj$matrixVals, parameterSet = parameterSet)

dataList0 <- readProjectData(fileLines = lines, progress = FALSE)

# add qfeatures

dataList <- add_qfeatures(dataList0, qfeatures = resultObj$qfeatures, fileAnnotation)

# *** gp: PR works until here to reproduce previous behavior with a simpler workflow

# use showcase for testing

# fileName <- system.file("extdata/showcase/Project_file_showcase_annotated.csv.gz", package = "MetFamily")
# project <- readClusterDataFromProjectFile(file = fileName)

project$dataFrameMS1Header$Annotation
dataList$dataFrameMS1Header$Annotation

Filtering data

We can filter the data to remove low quality data points

# todo...

PCA

PCA on MS1 in Figure \@ref(fig:pca).

fileName <- system.file("extdata/testdata/filterObj.Rdata", package = "MetFamily")
load(fileName) 

pca <- calculatePCA(dataList=dataList,
                         filterObj=filterObj, 
                         ms1AnalysisMethod="PCA (Principal Component Analysis)", 
                         scaling="None", 
                    logTransform=FALSE)

## Need to be global variables because they are not 
## passed to getPcaPerformanceIndicator() as parameters. 
pcaDimensionOne <<- 1
pcaDimensionTwo <<- 2

  resultObj <- calcPlotPCAscores(
    pcaObj = pca, 
    dataList = dataList, #project, 
    filterObj = filterObj,
    pcaDimensionOne = pcaDimensionOne, 
    pcaDimensionTwo = pcaDimensionTwo, 
    showScoresLabels = FALSE, 
    xInterval = NULL, 
    yInterval = NULL
)

  # undebug(calcPlotPCAloadings)
  # undebug(getPrecursorColors)

  resultObj <- calcPlotPCAloadings(
    pcaObj = pca, 
    dataList = dataList, #project, 
    filter = filterObj, 
    pcaDimensionOne = pcaDimensionOne, 
    pcaDimensionTwo = pcaDimensionTwo, 
    selectionFragmentPcaLoadingSet = NULL,
    selectionAnalysisPcaLoadingSet = NULL,
    selectionSearchPcaLoadingSet   = NULL,
    xInterval = NULL, 
    yInterval = NULL,
    loadingsLabels = "None", 
    showLoadingsAbundance = FALSE, 
    showLoadingsFeaturesAnnotated   = TRUE,
    showLoadingsFeaturesUnannotated = TRUE,
    showLoadingsFeaturesSelected    = TRUE,
    showLoadingsFeaturesUnselected  = TRUE
  )

HCA

HCA on MS2 in Figure \@ref(fig:hca).

if (FALSE) {
p <- calcPlotDendrogram_plotly(dataList=dataList, #project, 
                               filterObj=filterObj, 
                               clusterDataList=dataList, #project, 
  distanceMeasure = "Jaccard", 
  showClusterLabels=FALSE, 
  hcaPrecursorLabels="m/z / RT", 
  selectionFragmentTreeNodeSet = NULL, 
  selectionAnalysisTreeNodeSet = NULL, 
  selectionSearchTreeNodeSet = NULL, 
  selectedSelection, 
  heatmapContent, 
  heatmapOrdering, 
  heatmapProportion)
}

fileName <- system.file("extdata/testdata/clusterDataList.Rdata", package = "MetFamily")
load(fileName) 
fileName <- system.file("extdata/testdata/hcaFilter.Rdata", package = "MetFamily")
load(fileName) 

returnObj <- calcPlotDendrogram(dataList=dataList2, #project, 
                                filter=filter, 
                                clusterDataList=clusterDataList, 
                                annoPresentAnnotationsList = annoPresentAnnotationsList ,
                                annoPresentColorsList = annoPresentColorsList,
                                distanceMeasure="Jaccard (intensity-weighted)", 
                                selectionFragmentTreeNodeSet = NULL, 
                                selectionAnalysisTreeNodeSet = NULL, 
                                selectionSearchTreeNodeSet = NULL, 
                                showClusterLabels = TRUE, 
                                hcaPrecursorLabels = "m/z / RT", 
                                xInterval = c(1,219))

References {.unnumbered}

Appendix {.unnumbered}

Session info

sessionInfo()

ipb-halle/MetFamily documentation built on April 5, 2025, 5:05 p.m.