File I/O (pyOpenMS 3.6.0)

The bundled scripts use local files; they do not fetch vendor data or databases. Native tests cover tiny synthetic mzML, chromatogram, featureXML, consensusXML, idXML, FASTA, MGF and local TSV examples. Instrument conversion, mzIdentML/pepXML, TraML, and large-file throughput remain untested here.

mzML and indexed access

import pyopenms as ms
exp = ms.MSExperiment()
ms.MzMLFile().load("sample.mzML", exp)
print(exp.getNrSpectra(), exp.getNrChromatograms())
writer = ms.MzMLFile()
options = writer.getOptions()
options.setWriteIndex(True)
options.setCompression(True)
writer.setOptions(options)
writer.store("indexed.mzML", exp)

on_disk = ms.OnDiscMSExperiment()
if not on_disk.openFile("indexed.mzML"):
    raise ValueError("An indexed mzML is required")
for i in range(on_disk.getNrSpectra()):
    spec = on_disk.getSpectrum(i)
    # RT is seconds; get_peaks() yields m/z and intensity arrays.

IndexedMzMLFileLoader.load(filename, on_disk) is an alternative loader; it requires an OnDiscMSExperiment output argument. The loader itself is not a spectrum container. Preserve a known indexed input for lazy reading; converting an unindexed input via MSExperiment first still loads the full input in memory. CachedmzML.store(path, exp) and CachedmzML.load(path, cached) are available for local binary caches, which require their paired metadata files and disk space.

Format-specific I/O

proteins, peptides = ms.IdXMLFile().load("search.idXML")
ms.IdXMLFile().store("copy.idXML", proteins, peptides)
# The output-argument form also works:
proteins = []
peptides = ms.PeptideIdentificationList()
ms.IdXMLFile().load("search.idXML", proteins, peptides)

features = ms.FeatureMap()
ms.FeatureXMLFile().load("features.featureXML", features)
consensus = ms.ConsensusMap()
ms.ConsensusXMLFile().load("study.consensusXML", consensus)
entries = []
ms.FASTAFile().load("database.fasta", entries)

For untested formats, inspect installed signatures before adapting: help(ms.MzIdentMLFile.load), help(ms.PepXMLFile.load), help(ms.TraMLFile.load). pepXML is a general peptide-identification exchange format, not an X! Tandem-only format. An MSPFile.load call requires a PeptideIdentificationList and MSExperiment (or AnnotatedMSRun), not a list of spectra.

Metadata and loss

Use exp.getInstrument(), exp.getSample(), and exp.getSourceFiles() to inspect provenance. MSChromatogram.get_peaks() gives time in seconds and intensity; getNativeID() identifies stored traces, but a stored TIC need not be present or equal to the sum of the MS1 peaks retained after conversion.

convert_format.py preserves experiment metadata when filtering spectra, and retains synchronized peak arrays using MSSpectrum.select(indices). Its RT and intensity filters apply to spectra only. Chromatograms pass through unchanged unless --ms-level is selected, when they are omitted.

MGF exports tandem spectra and loses much mzML metadata, chromatograms, and other acquisition information; mzXML also cannot represent all mzML metadata. Round-trip counts alone do not establish lossless conversion. Keep the original file, conversion program/version/options, native IDs, charge/polarity, source metadata, and calibration state. Native vendor-reader availability in OpenMS 3.6 does not establish that any particular vendor file works in this wheel.

mzTab

Use the producing algorithm's populated MzTab/MzTabM with the corresponding writer. A title plus empty rows is not a complete report. When changing a copied metadata object, explicitly call mztab.setMetaData(metadata) before storing. Retain candidate multiplicity, adduct assumptions and database provenance.

Source: OpenMS release changes and installed pyOpenMS 3.6.0 binding signatures; the older mzML tutorial may show different output-argument conventions.