Adding inverse ion mobility information from an MGF to an MZML MS Convert spit out but without ion mobility.... stupid mzml format.

add_iims.py

#!/usr/bin/env python3
import re
import sys


def iter_iims(
    input_mgf: str,
    iim_pattern: str = r"IonMobility:(\d+\.\d+)",
):
    iim_pattern = re.compile(iim_pattern)
    with open(input_mgf, "r") as infile:
        for line in infile:
            matches = iim_pattern.findall(line)
            for match in matches:
                yield float(match)


def make_new_mzml(
    ion_mobility_values: str,
    input_mzml: str,
    output_mzml: str,
    pattern=re.compile(
        r'<cvParam cvRef="MS" accession="MS:1000016" name="scan start time" '
        r'value="\d+\.\d+" unitCvRef="UO" unitAccession="UO:0000010" unitName="second"\s*/>'
    ),
    line_for_iim: str = '              <cvParam cvRef="MS" accession="MS:1002815" name="inverse reduced ion mobility" value="{iim}" unitCvRef="MS" unitAccession="MS:1002814" unitName="volt-second per square centimeter"/>\n',
):
    i = 0
    with open(input_mzml, "r") as infile, open(output_mzml, "w") as outfile:
        for line in infile:
            line = line.rstrip("\n")
            outfile.write(line + "\n")
            if pattern.fullmatch(line.strip()):
                iim = ion_mobility_values[i]
                outfile.write(line_for_iim.format(iim=iim))
                i += 1
    assert i == len(
        ion_mobility_values
    ), "Number of IIMs to fill did not match the number of spectra."


if __name__ == "__main__":
    input_mgf, input_mzml, output_mzml = sys.argv[1:]

    ## test
    # input_mgf = "F17835_mgf_example_5spectra.mgf.txt"
    # input_mzml = "F17835_mzML_example_5spectra.mzML.txt"
    # output_mzml = "output.xml"

    ion_mobility_values = list(iter_iims(input_mgf))
    make_new_mzml(ion_mobility_values, input_mzml, output_mzml)