Research team
Expertise
I conduct bioinformatics research in mass spectrometry-based proteomics, aimed at supporting the study of protein composition in biological samples or tissues. I develop computational methods to improve the analysis of difficult-to-interpret mass spectrometry data. My focus is on the annotation of post-translational modifications and immunopeptides, which are particularly relevant in research on infectious diseases and cancer. In addition, I explore the ethical aspects of mass spectrometry data, especially as this technology is increasingly used in clinical settings.
Ground-Truth Benchmarking of Peptide Property Models for Indistinguishable Peptides.
Abstract
Mass spectrometry is a critical tool for the confident identification of complex peptide mixtures. However, a fundamental limitation remains: some peptides are analytically indistinguishable, exhibiting nearly identical retention times and fragmentation patterns. These indistinguishable peptides can lead to ambiguous or incorrect annotations, which is particularly problematic in fields such as immunopeptidomics, where a single misidentified peptide could misdirect therapeutic development. In this project, we will design and measure a carefully curated set of synthetic peptides to experimentally determine which peptides are indistinguishable by mass spectrometry. This benchmark dataset will then be used to evaluate whether state-of-the-art deep learning–based prediction tools accurately capture these analytical limitations. By comparing predicted and measured indistinguishability across a controlled set of positional isomers, we will assess how reliably computational models can serve as in silico proxies for real experiments, ultimately improving the interpretability, reliability, and confidence of peptide identifications in proteomics and immunopeptidomics.Researcher(s)
- Promoter: Adams Charlotte
Research team(s)
Project type(s)
- Research Project
Privacy in proteomics: Safeguarding personally identifiable information in clinical omics data.
Abstract
Advancements in mass spectrometry-based proteomics have revolutionized the study of complex biological systems, enabling the characterization of thousands of proteins from human samples in a single experiment. However, important questions have been raised on the potential ability to re-identify individuals via MS-based proteomics data. While genomic and transcriptomic data have been extensively studied for privacy risks, the privacy implications of proteomics data are mostly unknown. Inspired by facial recognition techniques, I propose a novel approach to identify privacy risks within clinical proteomics data. I then aim to mitigate these risks by developing an approach to de-identify the data, while preserving data utility. By addressing both the identification of privacy risks and the development of mitigation strategies, this project stands at the forefront of an emerging field. It tackles a problem that is poised to become critical in the near future as clinical proteomics data becomes more detailed and widely shared. The outcomes will provide a much-needed roadmap for secure and ethical data sharing in proteomics, ensuring that this field continues to drive scientific innovation while safeguarding individual privacy. This work has the potential to set new standards for privacy-conscious research in proteomics, establishing a foundation for ethically sustainable and impactful biomedical science.Researcher(s)
- Promoter: Bittremieux Wout
- Co-promoter: Laukens Kris
- Fellow: Adams Charlotte
Research team(s)
Project type(s)
- Research Project