Wout Bittremieux

The paper introduces mzQC, a standard for representing quality control metrics in mass spectrometry, enabling interoperable and machine-readable QC reporting.

CasanovoGUI provides a user-friendly desktop interface for Casanovo, a deep learning model for de novo peptide sequencing, simplifying setup and result interpretation.

This paper introduces SIMBA, a transformer-based model that learns interpretable structural similarity from tandem mass spectra to improve small molecule analog discovery.

This study demonstrates that MasSpec Pen technology can rapidly identify bacterial species in clinical isolates and specimens with high accuracy using molecular profiles of metabolites and lipids.

This study compares fermentation-induced molecular changes in cassava and cocoyam using metabolomics, revealing distinct catabolic and balanced metabolic responses.

The paper analyzes food metabolomics data to trace molecular signals from agricultural practices and environmental exposures, revealing contamination pathways and natural products.

TreeMS2 constructs molecular phylogenies from raw MS/MS data, revealing relationships based on molecular phenotype and identifying deviations from genetic-based trees.

pmultiqc generates metadata-aware QC reports for proteomics data using standardized metrics and integrates with existing workflows and cloud services.

The paper introduces an open-source platform for reference data-driven metabolomics analysis, demonstrating its ability to extract dietary patterns from stool metabolomics data.

The paper enhances ANN-SoLo for open modification searching by integrating predicted spectral libraries and improved vector representations, improving peptide identification in complex proteomics workflows.

The paper develops a deep learning model to predict MS2 identifiability using MS1 data and acquisition parameters, achieving 90% recovery of known identifiable spectra.

The paper improves Casanovo, a deep learning de novo peptidesequencer, by enhancing interpretability, generalizing its use in database search, speeding up performance, and adding tools for adoption and result interpretation.

ProteoBench provides a centralized platform for comparing proteomics data analysis workflows, enabling benchmarking and community collaboration.

SquiDBase provides a centralized repository for raw nanopore sequencing data from microbes, along with tools to process and expand reference datasets.

This study identifies temporal regulation of metabolic processes in Crocosphaera watsonii WH8501 using proteomics, revealing coordinated protein abundance patterns across diel cycles.

This paper introduces BAM, a method using biotransformation rules to annotate untargeted metabolomics spectra, achieving 24.2% correct structure assignments in a large molecular network.

The paper introduces MoSTERT and MoSTERT-2S, transformer-based models for zero-shot retention time prediction of peptides with unseen post-translational modifications, achieving state-of-the-art performance on an external dataset.

The paper develops a drug exposure resource from untargeted metabolomics data to identify medications without clinical records.

The paper modifies the Casanovo model to account for digestion enzyme bias but finds no significant improvement in sequencing accuracy.

This study uses semi-supervised learning on a large unlabeled dataset to improve cardiotoxicity prediction for three ion channels, identifying potential cardiac-active compounds.

Casanovo translates tandem mass spectra into peptide sequences using a transformer model, outperforming existing methods and improving analysis in immunopeptidomics and metaproteomics.

This paper discusses machine learning strategies to overcome data challenges in mass spectrometry-based proteomics, emphasizing the need for standardized datasets and larger data sizes to improve model performance.

The paper trains a deep learning model to predict fragment ion intensities, improving immunopeptide identification rates using timsTOF instruments.

The paper presents open-source software libraries for processing mzQC data in Python, R, and Java, enabling QC analysis and integration in mass spectrometry workflows.

The paper creates a suspect spectral library for untargeted metabolomics by propagating annotations from public MS/MS spectra to improve interpretation of unannotated spectra.

The paper introduces a cross-attention transformer encoder for multi-sequence inputs and evaluates it on peptide-MHCII binding prediction.

The paper shows that biased negative data in TCR-epitope binding prediction leads to poor performance in realistic scenarios.

This study uses interpretable deep learning to analyze TCR-epitope interactions and identify molecular binding patterns that influence TCR affinity.

This paper integrates semi-supervised machine learning into the ANN-SoLo tool to improve spectral library searching for open modification peptides, increasing identification rates.

The paper updates the spectrum_utils Python package to integrate mass spectrometry standards, enhance data processing, and unify visualization across proteomics and metabolomics applications.