Case Study: Molecular Signatures in Guillain-Barré Syndrome
Guillain-Barré Syndrome (GBS) is a rare autoimmune disorder where the body's immune system attacks peripheral nerves, leading to progressive muscle weakness and potential paralysis. With an incidence of only 1-2 cases per 100,000 people annually, understanding the molecular mechanisms underlying GBS remains challenging due to limited sample availability and the need for comprehensive, unbiased molecular profiling.
Traditional diagnostic approaches rely primarily on clinical presentation, cerebrospinal fluid analysis, and nerve conduction studies. However, circulating biomolecules in plasma may provide earlier, more specific signatures of disease progression and immune dysregulation—if we can rapidly identify and quantify them with confidence.
How do you comprehensively profile biomolecular alterations in rare neurological disorders to identify potential biomarkers for diagnosis, prognosis, or therapeutic monitoring?
AI-powered biomolecular-omics for biomarker discovery
Biomolecular-omics provides a functional readout of biological state, capturing the downstream effects of genetic, transcriptomic, and proteomic perturbations. Plasma biomolecules reflect systemic physiological changes and can serve as accessible biomarkers for disease monitoring.
Leveraging our proprietary Large Spectral Model (LSM), Pyxis rapidly and accurately converts raw mass spectrometry (MS) data into confident biomolecular identifications and absolute quantifications—transforming complex MS2 fragmentation patterns into actionable biological insights.
Application: Guillain-Barré Syndrome plasma metabolomics
We reanalyzed publicly available LC-MS/MS data from a clinical study comparing plasma samples from 30 GBS patients against 30 healthy controls. Plasma samples were analyzed using standard untargeted MS workflows, with MS2 fragmentation data acquired on a high-resolution mass spectrometer.
Raw MS data were processed through the Pyxis platform, which leveraged MS2 spectral matching powered by our LSM to deliver confident identifications and concentrations within minutes of data upload.
Distinct molecular signatures differentiate GBS from healthy controls
Pyxis identified and quantified biomolecules across multiple biochemical pathways, revealing clear metabolic dysregulation in GBS patients compared to healthy individuals.
Unsupervised clustering reveals disease-specific separation
UMAP dimensionality reduction analysis of the plasma immediately showed clear separation between GBS patients and healthy controls, with quality control samples clustering tightly together—demonstrating robust technical performance and genuine biological differences between cohorts.
Key observation: The biomolecular perturbations in GBS are substantial enough to drive cohort separation without any prior knowledge of disease status, suggesting strong, reproducible biomarker candidates.
Identifying highly significant metabolic alterations
Differential expression analysis revealed dozens of significantly altered biomolecules (p < 0.05, fold change > 2) between GBS patients and controls:
Most significantly upregulated biomolecules in GBS:
Glutamyl-valine (dipeptide, -log₁₀ p-value ~14): Dramatic elevation suggests altered protein catabolism or dipeptidyl peptidase activity
Hydroxy fatty acids (12,13-epoxy-9-octadecenoic acid, 9,12-octadecadiynoic acid): Multiple hydroxylated and oxidized lipid species indicate oxidative stress and lipid peroxidation
Long-chain fatty acids (tetranoeic acid derivatives, 12,15-tetranoeic acid): Alterations in fatty acid metabolism and inflammatory lipid mediators
Phospholipid-related metabolites (glycerol-3-phosphate, 1-O-hexadecyl-sn-glycero-3-phosphocholine): Changes in membrane lipid remodeling
Downregulated biomolecules:
Phenylacetylglutamine: Reduced levels may reflect altered gut microbiome-host co-metabolism or kidney function
Nucleotide precursors (cytidine diphosphate, guanosine monophosphate): Decreased purine/pyrimidine metabolism
Showing consistent patterns across individuals
Hierarchical clustering of the top differential biomolecules revealed consistent upregulation of oxidized lipids, hydroxy fatty acids, and dipeptides across most GBS patients, while healthy controls maintained uniformly lower levels. This consistency suggests these biomolecules represent robust, reproducible signatures of GBS pathophysiology rather than individual variability.
Biological interpretation: The predominance of oxidized and hydroxylated lipid species points to oxidative stress and inflammation as central features of GBS pathophysiology. Elevated dipeptides may reflect increased proteolysis or altered peptidase activity associated with autoimmune nerve damage.
Experimental details: Public data meets AI-powered reanalysis
Plasma samples from 30 GBS patients and 30 healthy controls were previously analyzed by LC-MS/MS in a published metabolomics study. Raw data files were reprocessed using the Pyxis platform.
Pyxis leveraged MS2 fragmentation spectra to deliver confident metabolite identifications through its Large Spectral Model, which has been trained on billions of mass spectra across diverse biological matrices and compound classes. Within 15 minutes of uploading data, Pyxis returned:
Confident identifications with MS2-based structural validation
Absolute quantifications for pathway-level interpretation
Statistical analysis and visualization for immediate biological insight
This reanalysis demonstrates how Pyxis can extract maximum value from existing datasets, identifying disease-specific signatures that inform on underlying pathophysiology.
Integrate Pyxis into your biomarker discovery pipeline
Pyxis transforms complex LC-MS/MS data into comprehensive molecular phenotypes for disease characterization. The combination of MS2-based structural confidence and rapid data processing enables:
Characterization of rare disease populations with small sample sizes
Identification of biomolecular signatures of drug response or disease progression
Validation biomarker candidates across multiple studies
Investigation of biomolecular mechanisms underlying neurological, autoimmune, and inflammatory disorders