input.json

{
  "mzML": {
    "schemaLocation": "http://psi.hupo.org/ms/mzml http://psidev.info/files/ms/mzML/xsd/mzML1.1.0.xsd",
    "id": "exampleData",
    "version": "1.1.0"
  },
  "cvList": {
    "cv": [
      {
        "id": "MS",
        "fullName": "Proteomics Standards Initiative Mass Spectrometry Ontology",
        "version": "3.79.0",
        "URI": "http://psidev.cvs.sourceforge.net/*checkout*/psidev/psi/psi-ms/mzML/controlledVocabulary/psi-ms.obo"
      }
    ],
    "cv.1": [
      {
        "id": "UO",
        "fullName": "Unit Ontology",
        "version": "12:10:2011",
        "URI": "http://obo.cvs.sourceforge.net/*checkout*/obo/obo/ontology/phenotype/unit.obo"
      }
    ],
    "count": "2"
  },
  "fileDescription": {
    "fileContent": {
      "cvParam": [
        {
          "cvRef": "MS",
          "accession": "MS:1000579",
          "name": "MS1 spectrum",
          "value": ""
        }
      ],
      "cvParam.1": [
        {
          "cvRef": "MS",
          "accession": "MS:1000580",
          "name": "MSn spectrum",
          "value": ""
        }
      ]
    },
    "sourceFileList": {
      "sourceFile": {
        "cvParam": [
          {
            "cvRef": "MS",
            "accession": "MS:1000768",
            "name": "Thermo nativeID format",
            "value": ""
          }
        ],
        "cvParam.1": [
          {
            "cvRef": "MS",
            "accession": "MS:1000563",
            "name": "Thermo RAW format",
            "value": ""
          }
        ],
        "cvParam.2": [
          {
            "cvRef": "MS",
            "accession": "MS:1000569",
            "name": "SHA-1",
            "value": "6679ba84f57e8f25a3b8ebecc806ecafc79492ec"
          }
        ],
        "id": "RAW1",
        "name": "UK001N1exoposb.raw",
        "location": "file:///"
      },
      "count": "1"
    }
  },
  "referenceableParamGroupList": {
    "referenceableParamGroup": {
      "cvParam": [
        {
          "cvRef": "MS",
          "accession": "MS:1002416",
          "name": "Orbitrap Fusion",
          "value": ""
        }
      ],
      "cvParam.1": [
        {
          "cvRef": "MS",
          "accession": "MS:1000529",
          "name": "instrument serial number",
          "value": "FSN10352"
        }
      ],
      "id": "CommonInstrumentParams"
    },
    "count": "1"
  },
  "softwareList": {
    "software": {
      "cvParam": [
        {
          "cvRef": "MS",
          "accession": "MS:1000532",
          "name": "Xcalibur",
          "value": ""
        }
      ],
      "id": "Xcalibur",
      "version": "1.1.982"
    },
    "software.1": {
      "cvParam": [
        {
          "cvRef": "MS",
          "accession": "MS:1000615",
          "name": "ProteoWizard software",
          "value": ""
        }
      ],
      "id": "pwiz",
      "version": "3.0.9205"
    },
    "count": "2"
  },
  "instrumentConfigurationList": {
    "instrumentConfiguration": {
      "referenceableParamGroupRef": [
        {
          "ref": "CommonInstrumentParams"
        }
      ],
      "componentList": {
        "source": {
          "cvParam": [
            {
              "cvRef": "MS",
              "accession": "MS:1000398",
              "name": "nanoelectrospray",
              "value": ""
            }
          ],
          "cvParam.1": [
            {
              "cvRef": "MS",
              "accession": "MS:1000485",
              "name": "nanospray inlet",
              "value": ""
            }
          ],
          "order": "1"
        },
        "analyzer": {
          "cvParam": [
            {
              "cvRef": "MS",
              "accession": "MS:1000081",
              "name": "quadrupole",
              "value": ""
            }
          ],
          "order": "2"
        },
        "analyzer.1": {
          "cvParam": [
            {
              "cvRef": "MS",
              "accession": "MS:1000484",
              "name": "orbitrap",
              "value": ""
            }
          ],
          "order": "3"
        },
        "detector": {
          "cvParam": [
            {
              "cvRef": "MS",
              "accession": "MS:1000624",
              "name": "inductive detector",
              "value": ""
            }
          ],
          "order": "4"
        },
        "count": "4"
      },
      "softwareRef": [
        {
          "ref": "Xcalibur"
        }
      ],
      "id": "IC1"
    },
    "instrumentConfiguration.1": {
      "referenceableParamGroupRef": [
        {
          "ref": "CommonInstrumentParams"
        }
      ],
      "componentList": {
        "source": {
          "cvParam": [
            {
              "cvRef": "MS",
              "accession": "MS:1000398",
              "name": "nanoelectrospray",
              "value": ""
            }
          ],
          "cvParam.1": [
            {
              "cvRef": "MS",
              "accession": "MS:1000485",
              "name": "nanospray inlet",
              "value": ""
            }
          ],
          "order": "1"
        },
        "analyzer": {
          "cvParam": [
            {
              "cvRef": "MS",
              "accession": "MS:1000081",
              "name": "quadrupole",
              "value": ""
            }
          ],
          "order": "2"
        },
        "analyzer.1": {
          "cvParam": [
            {
              "cvRef": "MS",
              "accession": "MS:1000083",
              "name": "radial ejection linear ion trap",
              "value": ""
            }
          ],
          "order": "3"
        },
        "detector": {
          "cvParam": [
            {
              "cvRef": "MS",
              "accession": "MS:1000253",
              "name": "electron multiplier",
              "value": ""
            }
          ],
          "order": "4"
        },
        "count": "4"
      },
      "softwareRef": [
        {
          "ref": "Xcalibur"
        }
      ],
      "id": "IC2"
    },
    "count": "2"
  },
  "dataProcessingList": {
    "dataProcessing": {
      "processingMethod": {
        "cvParam": [
          {
            "cvRef": "MS",
            "accession": "MS:1000544",
            "name": "Conversion to mzML",
            "value": ""
          }
        ],
        "order": "0",
        "softwareRef": "pwiz"
      },
      "id": "pwiz_Reader_Thermo_conversion"
    },
    "count": "1"
  },
  "run": {
    "id": "exampleData",
    "defaultInstrumentConfigurationRef": "IC1",
    "startTimeStamp": "2015-07-29 12:49:35Z",
    "defaultSourceFileRef": "RAW1",
    "scanPolarity": "positive"
  }
}

input.mzML

<?xml version="1.0" encoding="utf-8"?>
<indexedmzML xmlns="http://psi.hupo.org/ms/mzml" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://psi.hupo.org/ms/mzml http://psidev.info/files/ms/mzML/xsd/mzML1.1.2_idx.xsd">
  <mzML xmlns="http://psi.hupo.org/ms/mzml" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://psi.hupo.org/ms/mzml http://psidev.info/files/ms/mzML/xsd/mzML1.1.0.xsd" id="exampleData" version="1.1.0">
    <cvList count="2">
      <cv id="MS" fullName="Proteomics Standards Initiative Mass Spectrometry Ontology" version="3.79.0" URI="http://psidev.cvs.sourceforge.net/*checkout*/psidev/psi/psi-ms/mzML/controlledVocabulary/psi-ms.obo"/>
      <cv id="UO" fullName="Unit Ontology" version="12:10:2011" URI="http://obo.cvs.sourceforge.net/*checkout*/obo/obo/ontology/phenotype/unit.obo"/>
    </cvList>
    <fileDescription>
      <fileContent>
        <cvParam cvRef="MS" accession="MS:1000579" name="MS1 spectrum" value=""/>
        <cvParam cvRef="MS" accession="MS:1000580" name="MSn spectrum" value=""/>
      </fileContent>
      <sourceFileList count="1">
        <sourceFile id="RAW1" name="UK001N1exoposb.raw" location="file:///">
          <cvParam cvRef="MS" accession="MS:1000768" name="Thermo nativeID format" value=""/>
          <cvParam cvRef="MS" accession="MS:1000563" name="Thermo RAW format" value=""/>
          <cvParam cvRef="MS" accession="MS:1000569" name="SHA-1" value="6679ba84f57e8f25a3b8ebecc806ecafc79492ec"/>
        </sourceFile>
      </sourceFileList>
    </fileDescription>
    <referenceableParamGroupList count="1">
      <referenceableParamGroup id="CommonInstrumentParams">
        <cvParam cvRef="MS" accession="MS:1002416" name="Orbitrap Fusion" value=""/>
        <cvParam cvRef="MS" accession="MS:1000529" name="instrument serial number" value="FSN10352"/>
      </referenceableParamGroup>
    </referenceableParamGroupList>
    <softwareList count="2">
      <software id="Xcalibur" version="1.1.982">
        <cvParam cvRef="MS" accession="MS:1000532" name="Xcalibur" value=""/>
      </software>
      <software id="pwiz" version="3.0.9205">
        <cvParam cvRef="MS" accession="MS:1000615" name="ProteoWizard software" value=""/>
      </software>
    </softwareList>
    <instrumentConfigurationList count="2">
      <instrumentConfiguration id="IC1">
        <referenceableParamGroupRef ref="CommonInstrumentParams"/>
        <componentList count="4">
          <source order="1">
            <cvParam cvRef="MS" accession="MS:1000398" name="nanoelectrospray" value=""/>
            <cvParam cvRef="MS" accession="MS:1000485" name="nanospray inlet" value=""/>
          </source>
          <analyzer order="2">
            <cvParam cvRef="MS" accession="MS:1000081" name="quadrupole" value=""/>
          </analyzer>
          <analyzer order="3">
            <cvParam cvRef="MS" accession="MS:1000484" name="orbitrap" value=""/>
          </analyzer>
          <detector order="4">
            <cvParam cvRef="MS" accession="MS:1000624" name="inductive detector" value=""/>
          </detector>
        </componentList>
        <softwareRef ref="Xcalibur"/>
      </instrumentConfiguration>
      <instrumentConfiguration id="IC2">
        <referenceableParamGroupRef ref="CommonInstrumentParams"/>
        <componentList count="4">
          <source order="1">
            <cvParam cvRef="MS" accession="MS:1000398" name="nanoelectrospray" value=""/>
            <cvParam cvRef="MS" accession="MS:1000485" name="nanospray inlet" value=""/>
          </source>
          <analyzer order="2">
            <cvParam cvRef="MS" accession="MS:1000081" name="quadrupole" value=""/>
          </analyzer>
          <analyzer order="3">
            <cvParam cvRef="MS" accession="MS:1000083" name="radial ejection linear ion trap" value=""/>
          </analyzer>
          <detector order="4">
            <cvParam cvRef="MS" accession="MS:1000253" name="electron multiplier" value=""/>
          </detector>
        </componentList>
        <softwareRef ref="Xcalibur"/>
      </instrumentConfiguration>
    </instrumentConfigurationList>
    <dataProcessingList count="1">
      <dataProcessing id="pwiz_Reader_Thermo_conversion">
        <processingMethod order="0" softwareRef="pwiz">
          <cvParam cvRef="MS" accession="MS:1000544" name="Conversion to mzML" value=""/>
        </processingMethod>
      </dataProcessing>
    </dataProcessingList>
    <run id="exampleData" defaultInstrumentConfigurationRef="IC1" startTimeStamp="2015-07-29T12:49:35Z" defaultSourceFileRef="RAW1">
      <spectrumList count="36" defaultDataProcessingRef="pwiz_Reader_Thermo_conversion">
        <spectrum index="0" id="controllerType=0 controllerNumber=1 scan=3" defaultArrayLength="34573">
          <cvParam cvRef="MS" accession="MS:1000579" name="MS1 spectrum" value=""/>
          <cvParam cvRef="MS" accession="MS:1000511" name="ms level" value="1"/>
          <cvParam cvRef="MS" accession="MS:1000130" name="positive scan" value=""/>
          <cvParam cvRef="MS" accession="MS:1000128" name="profile spectrum" value=""/>
          <cvParam cvRef="MS" accession="MS:1000504" name="base peak m/z" value="432.239837646484" unitCvRef="MS" unitAccession="MS:1000040" unitName="m/z"/>
          <cvParam cvRef="MS" accession="MS:1000505" name="base peak intensity" value="3.8174825e06" unitCvRef="MS" unitAccession="MS:1000131" unitName="number of detector counts"/>
          <cvParam cvRef="MS" accession="MS:1000285" name="total ion current" value="2.0373532e07"/>
          <cvParam cvRef="MS" accession="MS:1000528" name="lowest observed m/z" value="148.509396632102" unitCvRef="MS" unitAccession="MS:1000040" unitName="m/z"/>
          <cvParam cvRef="MS" accession="MS:1000527" name="highest observed m/z" value="1616.112961912861" unitCvRef="MS" unitAccession="MS:1000040" unitName="m/z"/>
          <scanList count="1">
            <cvParam cvRef="MS" accession="MS:1000795" name="no combination" value=""/>
          </scanList>
        </spectrum>
      </spectrumList>
    </run>
  </mzML>
</indexedmzML>

input_processing.R

library(XML)
library(jsonlite)

source("R/file_metadata.R")
input_file <- "input.mzML"
input_list <- get_mzml_metadata(input_file)
input_json <- meta_export_json(input_list)

cat(input_json, file = "input.json")

mzML_meta_to_json.R

#' get mzML metadata
#'
#' @param mzml_file the mzML file to get metadata from
#'
#' @import XML
#' @export
get_mzml_metadata <- function(mzml_file){
  xml_doc <- xmlTreeParse(mzml_file, useInternalNodes = TRUE)
  ns <- xmlNamespaceDefinitions(xmlRoot(xml_doc), recursive = TRUE, simplify = TRUE)
  names(ns)[1] <- "d1"

  mz_metanodes <- getNodeSet(xml_doc, "/d1:indexedmzML/d1:mzML", ns)

  mz_meta <- list()
  tmp_attr <- unclass(xmlAttrs(mz_metanodes[[1]]))

  attr(tmp_attr, "namespaces") <- NULL
  mz_meta[["mzML"]][[".attrs"]] <- tmp_attr

  other_nodes_2_get <- c("cvList", "fileDescription",
                         "referenceableParamGroupList",
                         "softwareList",
                         "instrumentConfigurationList",
                         "dataProcessingList")

  other_nodes <- xmlChildren(mz_metanodes[[1]])
  other_list <- lapply(other_nodes, xmlToList)

  mz_meta <- c(mz_meta, other_list[other_nodes_2_get])

  mz_meta[["run"]][[".attrs"]] <- xmlAttrs(mz_metanodes[[1]][["run"]])

  mz_meta <- .remove_attrs(mz_meta)

  mz_meta_frame <- .to_data_frame(mz_meta)

  mz_meta_frame$run$scanPolarity <- .get_scan_polarity(other_list$run$spectrumList)

  mz_meta_frame$run$startTimeStamp <- gsub("T", " ", mz_meta_frame$run$startTimeStamp)

  mz_meta_frame
}

#' export metadata to json
#'
#' export the list metadata to a json string
#'
#' @param meta_list a list of metadata
#'
#' @importFrom jsonlite toJSON
#' @export
meta_export_json <- function(meta_list){
  toJSON(meta_list, pretty = TRUE, auto_unbox = TRUE)
}

#' transform to data frame
#'
#' @param in_list the list of xml nodes to work on
#'
.to_data_frame <- function(in_list){
  if (class(in_list) == "list") {
    out_list <- lapply(in_list, .to_data_frame)
  } else if (class(in_list) == "character") {
    if (!is.null(names(in_list))) {
      out_list <- as.data.frame(t(as.matrix(in_list)))
    } else {
      out_list <- in_list
    }

  }
  out_list
}

#' remove attributes
#'
#' removes a list entry called ".attrs" from a list, and makes them first level
#' partners
#'
#' @param in_list the list to work on
#'
.remove_attrs <- function(in_list){
  if (class(in_list) == "list") {
    out_list <- in_list
    list_names <- names(out_list)

    if (".attrs" %in% list_names) {
      tmp_attrs <- out_list[[".attrs"]]

      name_attrs <- names(tmp_attrs)

      if (sum(name_attrs %in% list_names) == 0) {
        for (i_name in name_attrs) {
          out_list[[i_name]] <- tmp_attrs[[i_name]]
        }
        out_list[[".attrs"]] <- NULL
      }
    } else {
      out_list <- lapply(out_list, .remove_attrs)
    }
    # still need to check the rest of the pieces of the list!
    out_list <- lapply(out_list, .remove_attrs)
  } else {
    out_list <- in_list
  }
  out_list
}

#' get_scan_mode
#'
#' takes a list from xmlToList for "run" and looks at whether all scans are positive, negative, or mixed
#'
#' @param spectrum_list the list of spectra
#'
.get_scan_polarity <- function(spectrum_list){
  spectrum_list[[".attrs"]] <- NULL
  scan_data <- lapply(spectrum_list, function(in_spectrum){
    cv_loc <- which(names(in_spectrum) %in% "cvParam")
    cv_data <- unlist(in_spectrum[cv_loc])
    scan_polarity <- grep("scan", cv_data, value = TRUE)

    scan_polarity
  })

  scan_polarity <- as.character(unique(scan_data))

  if ((length(scan_polarity) == 1) && (grepl("positive", scan_polarity))) {
    out_polarity <- "positive"
  } else if ((length(scan_polarity) == 1) && (grepl("negative", scan_polarity))) {
    out_polarity <- "negative"
  } else {
    out_polarity <- "mixed"
  }
  out_polarity
}