MLMathieu Lavallée-Adam

Mathieu Lavallée-Adam

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Mathieu Lavallée-Adam has a total of 265 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Desktop iDMS: A stand-alone application of intelligent differential mobility spectrometry (iDMS), a neural network that predicts optimal separation and compensation voltages for field asymmetric differential ion mobility spectrometry

    by Thao Nguyen-Tran, Emily Hashimoto-Roth, Xun-Xun Shi, Michael G. Organ, Theodore J. Perkins, Mathieu Lavallée-Adam, Steffany A. L. Bennett

    Summary: Mass spectrometry resolution of glycosphingolipids requires advanced separation methods. The virtually identical structures of glycolipid epimers cannot be discriminated by routine multiple reaction monitoring-high-performance liquid chromatography-electrospray ionization-tandem mass spectrometry (MRM-LC-ESI-MS/MS). Separation can be achieved using field asymmetric (differential) ion mobility mass spectrometry (FAIMS/DMS). A major bottleneck to MRM-LC-ESI-DMS-MS/MS method development lies in manually optimizing the compound-dependent parameters of separation voltage (SV) and compensation voltage (CoV). To address this challenge, we developed intelligent Differential Mobility Spectrometry (iDMS). iDMS is a supervised deep neural network that evaluates and predicts the optimal ion mobility SV and CoV parameters for any user-requested monoglycosphingolipid from a training set of measured signal intensities composed of 12 lipids (6 stereoisomer pairs). We initially introduced iDMS as a series of python scripts. Here, we share desktop applications of both iDMS and its pre-processing input data assembly modules (DMS Data Extraction Toolkit). Both software assemblies enable users to rapidly deploy a validated machine learning alternative to manual DMS optimization. The applications are user-friendly, operate without need for any technical operating environment, make training dataset assembly quick and simple, and return measured, normalized, and predicted ionograms at every user-interrogated SV, evaluation of separability at each SV and CoV combination, and prediction of optimal resolution parameters to accelerate users’ MRM-LC-ESI-DMS-MS/MS deployment. Availability and implementation: Desktop iDMS and DMS Data Extraction Toolkit are native desktop applications for macOS (Apple Silicon), freely available for download at https://www.neurolipidomics.com/platforms.html under the GNU v3.0 license. The source code is available from GitHub and Zenodo at https://github.com/Neurolipidomics/iDMS , https://zenodo.org/records/22214408 , https://github.com/Neurolipidomics/DMSDataExtractionToolkit , and https://ze-nodo.org/records/22777403 . Benchmarking metrics are available at https://www.biorxiv.org/content/10.64898/2026.08.26.747394v1 . Contact: ldomic@uottawa.ca Supplementary Information: Supplementary material is available at Bioinformatics online.

  • TargetSeeker-MS: A Bayesian Inference Approach for Drug–Target Discovery Using Protein Fractionation Coupled to Mass Spectrometry

    by Mathieu Lavallée-Adam, Alexander Pelletier, Jolene K. Diedrich, Antonio F. M. Pinto, Salvador Martínez-Bartolomé, Michael Petrascheck, James J. Moresco, John R. Yates

    To understand the mechanism of action of a drug and assess its clinical usefulness and viability, it is imperative that its affinity for its putative targets is determined. When coupled to mass spectrometry (MS), energetics-based protein separation (EBPS) techniques, such as a thermal shift assay, have shown great potential to identify the targets of a drug on a proteome scale. Nevertheless, the computational analyses assessing the confidence of drug-target predictions made by these methods have remained tightly tied to the protocol under which the data were produced. To identify drug targets in data sets produced using different EBPS-MS techniques, we have developed a novel flexible Bayesian inference approach named TargetSeeker-MS. We showed that TargetSeeker-MS identifies known and novel drug targets in Caenorhabditis elegans and HEK 293 samples treated with the fungicide benomyl. We also demonstrated that TargetSeeker-MS' drug-target identifications are reproducible in C. elegans samples that were processed using two different EBPS techniques (thermal shift assay and a differential precipitation of proteins, named DiffPOP). In addition, we validated a novel benomyl target by measuring its altered enzymatic activity upon drug treatment in vitro. TargetSeeker-MS, which is available as a web server (https://targetseeker.scripps.edu/), allows for the rapid, versatile, and confident identification of targets of a drug on a proteome scale, thereby providing a better understanding of its mechanisms and facilitating the evaluation of its clinical viability.

  • MAGPIE: A Machine Learning Approach to Decipher Protein–Protein Interactions in Human Plasma

    by Emily Hashimoto-Roth, Diane Forget, Vanessa P. Gaspar, Steffany A. L. Bennett, Marie-Soleil Gauthier, Benoit Coulombe, Mathieu Lavallée-Adam

    Immunoprecipitation coupled to tandem mass spectrometry (IP-MS/MS) methods is often used to identify protein-protein interactions (PPIs). While these approaches are prone to false positive identifications through contamination and antibody nonspecific binding, their results can be filtered using negative controls and computational modeling. However, such filtering does not effectively detect false-positive interactions when IP-MS/MS is performed on human plasma samples. Therein, proteins cannot be overexpressed or inhibited, and existing modeling algorithms are not adapted for execution without such controls. Hence, we introduce MAGPIE, a novel machine learning-based approach for identifying PPIs in human plasma using IP-MS/MS, which leverages negative controls that include antibodies targeting proteins not expected to be present in human plasma. A set of negative controls used for false positive interaction modeling is first constructed. MAGPIE then assesses the reliability of PPIs detected in IP-MS/MS experiments using antibodies that target known plasma proteins. When applied to five IP-MS/MS experiments as a proof of concept, our algorithm identified 68 PPIs with an FDR of 20.77%. MAGPIE significantly outperformed a state-of-the-art PPI discovery tool and identified known and predicted PPIs. Our approach provides an unprecedented ability to detect human plasma PPIs, which enables a better understanding of biological processes in plasma.


  • Open Access

    Machine learning-enhanced immunopeptidomics applied to T-cell epitope discovery for COVID-19 vaccines

    by Kevin A. Kovalchik, David J. Hamelin, Peter Kubiniok, Benoîte Bourdin, Fatima Mostefai, Raphaël Poujol, Bastien Paré, Shawn M. Simpson, John Sidney, Éric Bonneil, et al.

    Next-generation T-cell-directed vaccines for COVID-19 focus on establishing lasting T-cell immunity against current and emerging SARS-CoV-2 variants. Precise identification of conserved T-cell epitopes is critical for designing effective vaccines. Here we introduce a comprehensive computational framework incorporating a machine learning algorithm—MHCvalidator—to enhance mass spectrometry-based immunopeptidomics sensitivity. MHCvalidator identifies unique T-cell epitopes presented by the B7 supertype, including an epitope from a + 1-frameshift in a truncated Spike antigen, supported by ribosome profiling. Analysis of 100,512 COVID-19 patient proteomes shows Spike antigen truncation in 0.85% of cases, revealing frameshifted viral antigens at the population level. Our EpiTrack pipeline tracks global mutations of MHCvalidator-identified CD8 + T-cell epitopes from the BNT162b4 vaccine. While most vaccine epitopes remain globally conserved, an immunodominant A*01-associated epitope mutates in Delta and Omicron variants. This work highlights SARS-CoV-2 antigenic features and emphasizes the importance of continuous adaptation in T-cell vaccine development.


  • Open Access

    Integrating Machine Learning-Enhanced Immunopeptidomics and SARS-CoV-2 Population-Scale Analyses Unveils Novel Antigenic Features for Next-Generation COVID-19 Vaccines

    by Etienne Caron, Kevin Kovalchik, David Hamelin, Peter Kubiniok, Benoite Bourdin, Fatima Mostefai, Raphaël Poujol, Bastien Paré, Shawn Simpson, John Sidney, et al.

    Next-generation T-cell-directed vaccines for COVID-19 aim to induce durable T-cell immunity against circulating and future hypermutated SARS-CoV-2 variants. Mass Spectrometry (MS)-based immunopeptidomics holds promise for guiding vaccine design, but computational challenges impede the precise and unbiased identification of conserved T-cell epitopes crucial for vaccines against rapidly mutating viruses. We introduce a computational framework and analysis platform integrating a novel machine learning algorithm, immunopeptidomics, intra-host data, epitope immunogenicity, and geo-temporal CD8+ T-cell epitope conservation analyses. Central to our approach is MHCvalidator, a novel artificial neural network algorithm enhancing MS-based immunopeptidomics sensitivity by modeling antigen presentation and sequence features. MHCvalidator identified a novel nonconventional SARS-CoV-2 T-cell epitope presented by B7 supertype molecules, originating from a +1-frameshift in a truncated Spike (S) antigen, supported by ribo-seq data. Intra-host analysis of SARS-CoV-2 proteomes from ~100,000 COVID-19 patients revealed a prevalent S antigen truncation in ~51% of cases, exposing a rich source of frameshifted viral antigens. Our framework includes EpiTrack, a new computational pipeline tracking global mutational dynamics of MHCvalidator-identified SARS-CoV-2 CD8+ epitopes from vaccine BNT162b4. While most vaccine-encoded CD8+ epitopes exhibit global conservation from January 2020 to October 2023, a highly immunodominant A*01-associated epitope, especially in hospitalized patients, undergoes substantial mutations in Delta and Omicron variants. Our approach unveils unprecedented SARS-CoV-2 T-cell epitopes, elucidates novel antigenic features, and underscores mutational dynamics of vaccine-relevant epitopes. The analysis platform is applicable to any viruses, and underscores the need for continual vigilance in T-cell vaccine development against the evolving landscape of hypermutating SARS-CoV-2 variants.


  • METAbolomics data Balancing with Over-sampling Algorithms (META-BOA): an online resource for addressing class imbalance

    by Emily Hashimoto-Roth, Anuradha Surendra, Mathieu Lavallée-Adam, Steffany A L Bennett, Miroslava Čuperlović-Culf

    Class imbalance, or unequal sample sizes between classes, is an increasing concern in machine learning for metabolomic and lipidomic data mining, which can result in overfitting for the over-represented class. Numerous methods have been developed for handling class imbalance, but they are not readily accessible to users with limited computational experience. Moreover, there is no resource that enables users to easily evaluate the effect of different over-sampling algorithms. METAbolomics data Balancing with Over-sampling Algorithms (META-BOA) is a web-based application that enables users to select between four different methods for class balancing, followed by data visualization and classification of the sample to observe the augmentation effects. META-BOA outputs a newly balanced dataset, generating additional samples in the minority class, according to the user’s choice of Synthetic Minority Over-sampling Technique (SMOTE), Borderline-SMOTE (BSMOTE), Adaptive Synthetic (ADASYN) or Random Over-Sampling Examples (ROSE). To present the effect of over-sampling on the data META-BOA further displays both principal component analysis and t-distributed stochastic neighbor embedding visualization of data pre- and post-over-sampling. Random forest classification is utilized to compare sample classification in both the original and balanced datasets, enabling users to select the most appropriate method for their further analyses. META-BOA is available at https://complimet.ca/meta-boa. Supplementary data are available at Bioinformatics online.


  • Novel Bioinformatics Strategies Driving Dynamic Metaproteomic Studies

    by Caitlin M. A. Simopoulos, Daniel Figeys, Mathieu Lavallée-Adam

    Constant improvements in mass spectrometry technologies and laboratory workflows have enabled the proteomics investigation of biological samples of growing complexity. Microbiomes represent such complex samples for which metaproteomics analyses are becoming increasingly popular. Metaproteomics experimental procedures create large amounts of data from which biologically relevant signal must be efficiently extracted to draw meaningful conclusions. Such a data processing requires appropriate bioinformatics tools specifically developed for, or capable of handling metaproteomics data. In this chapter, we outline current and novel tools that can perform the most commonly used steps in the analysis of cutting-edge metaproteomics data, such as peptide and protein identification and quantification, as well as data normalization, imputation, mining, and visualization. We also provide details about the experimental setups in which these tools should be used.


  • Open Access

    Proximity Interactome Map of the Vac14–Fig4 Complex Using BioID

    by Shirley Qiu, Mathieu Lavallée-Adam, Marceline Côté

    Conversion between phosphatidylinositol-3-phosphate and phosphatidylinositol-3,5-bisphosphate on endosomal membranes is critical for the maturation of early endosomes to late endosomes/lysosomes and is regulated by the PIKfyve–Vac14–Fig4 complex. Despite the importance of this complex for endosomal homeostasis and vesicular trafficking, there is little known about how its activity is regulated or how it interacts with other cellular proteins. Here, we screened for the cellular interactome of Vac14 and Fig4 using proximity-dependent biotin labeling (BioID). After independently screening the interactomes of Vac14 and Fig4, we identified 89 high-confidence protein hits shared by both proteins. Network analysis of these hits revealed pathways with known involvement of the PIKfyve–Vac14–Fig4 complex, including vesicular organization and PI3K/Akt signaling, as well as novel pathways including cell cycle and mitochondrial regulation. We also identified subunits of coatomer complex I (COPI), a Golgi-associated complex with an emerging role in endosomal dynamics. Using proximity ligation assays, we validated the interaction between Vac14 and COPI subunit COPB1 and between Vac14 and Arf1, a GTPase required for COPI assembly. In summary, this study used BioID to comprehensively map the Vac14–Fig4 interactome, revealing potential roles for these proteins in diverse cellular processes and pathways, including preliminary evidence of an interaction between Vac14 and COPI. Data are available via ProteomeXchange with the identifier PXD027917.


  • Identification of novel lipid droplet factors that regulate autophagy and cholesterol efflux in macrophage foam cells

    by S. Robichaud, G. Fairman, V. Vijithakumar, E. Mak, D. Cook, A. Pelletier, B. Vanderhyden, D. Figeys, M. Lavallée-Adam, K. Baetz, et al.


  • Open Access

    PIGNON: a protein–protein interaction-guided functional enrichment analysis for quantitative proteomics

    by Rachel Nadeau, Anastasiia Byvsheva, Mathieu Lavallée-Adam

    Background: Quantitative proteomics studies are often used to detect proteins that are differentially expressed across different experimental conditions. Functional enrichment analyses are then typically used to detect annotations, such as biological processes that are significantly enriched among such differentially expressed proteins to provide insights into the molecular impacts of the studied conditions. While common, this analytical pipeline often heavily relies on arbitrary thresholds of significance. However, a functional annotation may be dysregulated in a given experimental condition, while none, or very few of its proteins may be individually considered to be significantly differentially expressed. Such an annotation would therefore be missed by standard approaches. Results: Herein, we propose a novel graph theory-based method, PIGNON, for the detection of differentially expressed functional annotations in different conditions. PIGNON does not assess the statistical significance of the differential expression of individual proteins, but rather maps protein differential expression levels onto a protein–protein interaction network and measures the clustering of proteins from a given functional annotation within the network. This process allows the detection of functional annotations for which the proteins are differentially expressed and grouped in the network. A Monte-Carlo sampling approach is used to assess the clustering significance of proteins in an expression-weighted network. When applied to a quantitative proteomics analysis of different molecular subtypes of breast cancer, PIGNON detects Gene Ontology terms that are both significantly clustered in a protein–protein interaction network and differentially expressed across different breast cancer subtypes. PIGNON identified functional annotations that are dysregulated and clustered within the network between the HER2+, triple negative and hormone receptor positive subtypes. We show that PIGNON’s results are complementary to those of state-of-the-art functional enrichment analyses and that it highlights functional annotations missed by standard approaches. Furthermore, PIGNON detects functional annotations that have been previously associated with specific breast cancer subtypes. Conclusion: PIGNON provides an alternative to functional enrichment analyses and a more comprehensive characterization of quantitative datasets. Hence, it contributes to yielding a better understanding of dysregulated functions and processes in biological samples under different experimental conditions.