The University of Nebraska Medical Center (UNMC) is spotlighting its Multiomics Mass Spectrometry Core facility as advancements in high-throughput mass spectrometry (MS) drive demand for standardized data formats, according to a recent announcement.
For years, proteomics research was hampered by proprietary data formats tied to specific instrument manufacturers. These “vendor lock-in” systems hindered large-scale data sharing and meta-analyses, slowing scientific progress. The Proteomics Standards Initiative (PSI) of the Human Proteome Organization (HUPO) addressed this issue by developing open-source, XML-based standards for storing, sharing, and re-analyzing MS data.
The transition to open standards, a decade-long process, prioritizes transparency, and reproducibility. While manufacturer-specific formats offer speed and storage efficiency, they lack the openness needed for rigorous peer review and cross-platform validation. Standardizing the recording of mass-to-charge ratios, intensities, and metadata has laid the groundwork for “Big Data” proteomics, enabling global repositories and cloud-based analysis.
The mzML format serves as the foundational standard for raw mass spectrometry data. Developed from earlier formats mzData and mzXML in 2008, mzML provides a unified structure for metadata and mass spectral peaks, regardless of the instrument used. The format’s hierarchical structure, built using eXtensible Markup Language (XML), comprehensively describes instrument hardware, software settings, data acquisition parameters, and scan data. Binary data is encoded using Base64 to ensure compatibility with XML, and compression algorithms like MS-Numpress mitigate file size increases without sacrificing precision.
Key components of an mzML file include controlled vocabularies – standardized terms ensuring machine-readable metadata – data processing history for auditing data transformations, and detailed scan settings for tandem MS experiments.
While mzML handles raw data, mzIdentML reports protein and peptide identification results. This format captures peptide-spectrum matches (PSMs), protein groups, and associated confidence scores. A key advantage of mzIdentML is its ability to address the “protein inference” problem, accurately representing protein groups and the evidence supporting each member, while recording search engine details and database versions.
The quantification of proteins requires additional standardization, addressed by mzQuantML. This format represents data from various quantitative workflows, including label-free quantification and metabolic labeling, focusing on the evidence for quantification – peak areas, elution profiles, or reporter ion intensities – allowing for re-examination of underlying data. For simpler summary reports, mzTab, a tab-delimited text format, provides a human-readable and machine-parsable alternative, ideal for submission to public repositories like PRIDE.
The adoption of these standards is linked to the FAIR principles – Findability, Accessibility, Interoperability, and Reusability. Consortia like ProteomeXchange mandate PSI standards for data deposition, enabling integration into larger meta-analyses and automated validation of uploaded data. This supports projects like ProteomicsDB and the Human Protein Atlas, which aggregate data from thousands of experiments.
Despite widespread adoption, challenges remain. File size remains a significant limitation, with XML-based formats being inherently verbose. Conversion processes require constant updates to align with new instrument firmware, and potential information loss during data centroiding can impact advanced post-processing algorithms. Panome Bio recently added global phosphoproteomics to its portfolio of multi-omic discovery solutions, further highlighting the growing demand for integrated workflows.
The increasing focus on multiomics, integrating data from various sources, is further driving the need for standardized formats, as highlighted by recent discussions on achieving true multiomic workflows in spatial biology and disease research.
