In lifs-tools/rmzTab-m: R Package Client for mzTab-M Reference Implementation and Validation API

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)

library(rmzTabM)
library(kableExtra)

Reading mzTab-M files

You can read an mzTab-M file in tab separated format into a data frame structure as follows:

testfile <- system.file("testdata", c("lipidomics-example.mzTab"), package="rmzTabM")
mzTabTable <- readMzTab(testfile)
kable(head(mzTabTable[,1:3])) %>% kable_styling(bootstrap_options = c("striped", "hover", "condensed", font_size = 7)) %>% scroll_box(width = "800px", height = "200px")

You can extract the individual section tables from this one as follows:

mtdTable <- extractMetadata(mzTabTable)
smlTable <- extractSmallMoleculeSummary(mzTabTable)
smfTable <- extractSmallMoleculeFeatures(mzTabTable)
smeTable <- extractSmallMoleculeEvidence(mzTabTable)

knitr::kable(head(smlTable)) %>% kable_styling(bootstrap_options = c("striped", "hover", "condensed", font_size = 7)) %>% scroll_box(width = "800px", height = "200px")

knitr::kable(head(smfTable)) %>% kable_styling(bootstrap_options = c("striped", "hover", "condensed", font_size = 7)) %>% scroll_box(width = "800px", height = "200px")

knitr::kable(head(smeTable)) %>% kable_styling(bootstrap_options = c("striped", "hover", "condensed", font_size = 7)) %>% scroll_box(width = "800px", height = "200px")

To turn these tables into objects, use the R6 class constructor method new():

mzTabObject <- MzTab$new()$fromDataFrame(mzTabTable)

You can then access sections like Metadata as object members:

# this is an R6 object
metadata <- mzTabObject$metadata
# these are lists
smallMoleculeSummaryEntries <- mzTabObject$smallMoleculeSummary
smallMoleculeFeatureEntries <- mzTabObject$smallMoleculeFeature
smallMoleculeEvidenceEntries <- mzTabObject$smallMoleculeEvidence

Extracting values is possible from either representation, depending on whether you prefer a tabular style or an object oriented style, however, there may be type differences:

# this is the SmallMoleculeSummary list first entry id
smallMoleculeSummaryEntries[[1]]$sml_id

# and this is the same with the data frame
as.numeric(smlTable$SML_ID)

Writing mzTab-M files

If you have an mzTab-M data frame, you can write it as follows:

utils::write.table(
    mzTabTable,
    file = file.path(tempdir(check=TRUE), "mzTabWrite1.mztab"),
    row.names = FALSE,
    col.names = FALSE,
    quote = FALSE,
    sep = "\t",
    na = "",
    fileEncoding = "UTF8"
  )

For an MzTab object, you can write to the tab separated format:

writeMzTab(mzTabObject, file.path(tempdir(check=TRUE), "mzTabWrite2.mztab"))

Or to JSON format:

writeMzTabJSON(mzTabObject, file.path(tempdir(check=TRUE), "mzTabWrite3.mztab.json"))

Validating mzTab-M files

To validate an mzTab-M file, you can access the mzTab Validator web application at https://apps.lifs-tools.org/mztabvalidator

You can set the validationLevel to one of info, warning or error. If you enable semanticValidation, CV parameters present in your file will be checked against the default recommended mapping file.

In order to validate an mzTab-M file without needing to parse it locally, use the following call, which should return an empty list:

  validatePlainFile <- system.file("testdata", c("lipidomics-example.mzTab"),package="rmzTabM")
  mzTabString <- readChar(validatePlainFile, file.info(testfile)$size)
  validationMessages2 <- validateMzTab(
    mzTabString,
    validationMode = "plain",
    validationLevel = "info",
    maxErrors = 100,
    semanticValidation = FALSE
  )
  if(length(validationMessages2)==0) {
    print("No validation messages")
  } else {
    validationMessages2
  }

Alternatively, to run the validation with semantic checks of the used CV parameters against the default mapping file, which will give you hints on how to improve your file:

  validationMessages2 <- validateMzTab(
    mzTabString,
    validationMode = "plain",
    validationLevel = "info",
    maxErrors = 100,
    semanticValidation = TRUE
  )
  dfList<-lapply(validationMessages2, function(x) { data.frame("Category"=x$category, "Message"=x$code) })
  vmdf <- do.call("rbind", dfList)
  knitr::kable(vmdf) %>% kable_styling(bootstrap_options = c("striped", "hover", "condensed", font_size = 7)) %>% scroll_box(width = "800px", height = "1000px")

lifs-tools/rmzTab-m documentation built on Jan. 26, 2023, 4:45 p.m.