Spatial Proteomics & MALDI-MSI
Advancing digital pathology and biomarker discovery through Matrix-Assisted Laser Desorption Ionization Mass Spectrometry Imaging (MALDI-MSI) and machine learning.
Spatial Proteomics combined with MALDI-MSI offers an unprecedented, label-free approach to mapping molecular distributions directly within intact tissue architectures.
Our research leverages high-resolution mass spectrometry imaging to extract complex molecular fingerprints from tissue samples. By coupling these massive proteomic datasets with advanced computational pathology and machine learning algorithms, we identify novel biomarkers, distinguish challenging tumor histologies, and uncover the distinct pathogenic pathways driving disease progression.
The Power of Mass Spectrometry Imaging
Traditional histology relies heavily on morphology, which can sometimes result in ambiguous diagnoses for borderline or highly heterogeneous lesions. MALDI-MSI bridges the gap between molecular biology and morphology by maintaining the spatial context of proteins, peptides, and metabolites.
The core advantages of our approach include:
- Untargeted Discovery: Unlike immunohistochemistry, MALDI-MSI does not require a priori knowledge of the target. We can analyze hundreds of molecules simultaneously in a single tissue section.
- Morphological Correlation: Extracted mass spectra are directly mapped to specific regions of interest (ROIs) annotated by expert pathologists, linking pure molecular data to histological ground truth.
- Diagnostic Precision: By identifying unique spatial proteomic signatures, we can stratify tumor subtypes, assess mutational statuses (e.g., NRAS mutations), and minimize diagnostic uncertainty.
Our Research Focus & Methodologies
Across multiple ongoing studies, our group develops and applies robust statistical and machine learning frameworks to handle the high dimensionality of MALDI-MSI data.
- Feature Selection & Classification: We employ supervised and unsupervised machine learning models to sift through thousands of signals. This allows us to isolate the most impactful proteomic features responsible for disease classification.
- Translational Biomarker Identification: Once key spatial signals are isolated, we utilize nanoscale liquid chromatography electrospray ionization tandem mass spectrometry (nLC-ESI-MS/MS) to confidently identify the underlying proteins, translating raw signals into actionable biological insights.
- Multi-Tissue Applications: While we have extensively applied these pipelines to indeterminate thyroid neoplasms, our computational workflows are designed to be scalable and adaptable to a wide range of oncological and pathological challenges.
Research Highlights
High-Throughput MSI
Utilizing state-of-the-art MALDI-TOF/TOF systems to extract spatially resolved molecular features directly from tissue microarrays and whole slides.
AI-Driven Discovery
Applying dimensionality reduction and classification algorithms to decode complex molecular spectra into reliable diagnostic models.
Clinical Translation
Bridging the gap between untargeted omics and digital pathology to provide reliable ancillary tools for personalized medicine.
Core Publication
2026
- Improving the Annotation for Spatial Proteomics: A Computational Approach to Enhance Molecular Characterization of Thyroid NodulesJournal of Proteome Research, 2026
- Spatial Proteomics Uncovers Region-Specific Proteomic Signatures in Autoimmune Liver DiseasesDigestive and Liver Disease, 2026
2025
- Well begun Is half done: The impact of pre-processing in MALDI mass spectrometry imaging analysis applied to a case study of thyroid nodulesStats, 2025
- Machine Learning ensemble algorithms for classification of thyroid nodules through proteomics: Extending the method of Shapley values from binary to multi-class tasksStats, 2025
- Biomarker identification through spatial proteomics for the characterization of indeterminate thyroid nodulesEndocrine, 2025
2024
-
- Spatial resolution of renal amyloid deposits through MALDI-MSI: a combined digital and molecular approach to monoclonal gammopathiesJournal of Clinical Pathology, 2024
2023
- Unsupervised neural networks as a support tool for pathology diagnosis in MALDI-MSI experiments: A case study on thyroid biopsiesExpert Systems with Applications, 2023
- Segmenting Brain MALDI-MSI Data Under Separate ExchangeabilityIn Interational Conference on Bayesian Young Statistician Meeting, 2023
2021
- Lipidomic typing of colorectal cancer tissue containing tumour-infiltrating lymphocytes by MALDI mass spectrometry imagingMetabolites, 2021
2020
- Molecular trait of follicular-patterned thyroid neoplasms defined by MALDI-imagingBiochimica et Biophysica Acta (BBA)-Proteins and Proteomics, 2020
- Analysis of Hashimoto’s thyroiditis on fine needle aspiration samples by MALDI-ImagingBiochimica et Biophysica Acta (BBA)-Proteins and Proteomics, 2020
2019
- Molecular signatures of medullary thyroid carcinoma by matrix-assisted laser desorption/ionisation mass spectrometry imagingJournal of proteomics, 2019
- Feasibility study for the MALDI-MSI analysis of thyroid fine needle aspiration biopsies: evaluating the morphological and proteomic stability over timePROTEOMICS–Clinical Applications, 2019