Exploring geographical differences in IgE response through network and manifold analyses

• Published in: Journal of allergy and clinical immunology
• Main Authors: Cucco, Alex, Pearce, Neil, Simpson, Angela, Pembrey, Lucy, Mpairwe, Harriet, Figueiredo, Camila A, Cooper, Philip J, Douwes, Jeroen, Brooks, Collin, Adcock, Ian M, Kermani, Nazanin Zounemat, Roberts, Graham Dm, Murray, Clare S, Custovic, Adnan, Fontanella, Sara
• Format: Journal Article
• Language: English
• Published: 30.06.2025
• ISSN: 1097-6825; 1097-6825
• DOI: 10.1016/j.jaci.2025.05.032

Component-resolved diagnostics (CRD) allows detailed assessment of IgE-sensitization to multiple allergenic molecules (c-sIgEs) and may be useful for asthma diagnosis. However, to effectively use CRD across diverse settings, it is crucial to account for geographical differences.BACKGROUNDComponent-resolved diagnostics (CRD) allows detailed assessment of IgE-sensitization to multiple allergenic molecules (c-sIgEs) and may be useful for asthma diagnosis. However, to effectively use CRD across diverse settings, it is crucial to account for geographical differences.To investigate spatial determinants of c-sIgE networks to facilitate development of diagnostic algorithms applicable globally.OBJECTIVETo investigate spatial determinants of c-sIgE networks to facilitate development of diagnostic algorithms applicable globally.We used multiplex CRD array to measure c-sIgE to 112 proteins in an international collaboration of several studies: WASP consortium (UK, New Zealand, Brazil, Ecuador and Uganda), U-BIOPRED consortium (7 European countries) and a UK population-based birth cohort (MAAS). We employed hierarchical clustering to ascertain sensitization and c-sIgE clusters across diverse populations. Cross-country comparisons focused on a common subset of 18 c-sIgEs, using co-occurrence network representations projected into a low-dimensional Euclidean space to emphasize similarities, and spectral clustering to identify characteristic c-sIgEs clusters foremost characterizing the differences across geographical location. We then investigated sensitization networks across diverse regions in relation to asthma severity.METHODSWe used multiplex CRD array to measure c-sIgE to 112 proteins in an international collaboration of several studies: WASP consortium (UK, New Zealand, Brazil, Ecuador and Uganda), U-BIOPRED consortium (7 European countries) and a UK population-based birth cohort (MAAS). We employed hierarchical clustering to ascertain sensitization and c-sIgE clusters across diverse populations. Cross-country comparisons focused on a common subset of 18 c-sIgEs, using co-occurrence network representations projected into a low-dimensional Euclidean space to emphasize similarities, and spectral clustering to identify characteristic c-sIgEs clusters foremost characterizing the differences across geographical location. We then investigated sensitization networks across diverse regions in relation to asthma severity.Hierarchical clustering revealed that despite variations in the prevalence of specific c-sIgE, sensitization profiles shared similarities across regions. A unique component cluster with only alpha-gal sensitization was noted in Uganda and Ecuador. For 18 c-sIgEs shared across study populations, the structure of response allowed for differentiation between different geographical areas and study design, revealing 3 clusters: (1) Uganda, Ecuador, and Brazil; (2) U-BIOPRED children and adults; (3) New Zealand, UK and MAAS. Spectral clustering identified differences between clusters, with 5 c-sIgEs groups solution characterizing Cluster 1. Finally, we observed almost parallel and constant shifts between severe and non-severe asthma subjects in each country.RESULTSHierarchical clustering revealed that despite variations in the prevalence of specific c-sIgE, sensitization profiles shared similarities across regions. A unique component cluster with only alpha-gal sensitization was noted in Uganda and Ecuador. For 18 c-sIgEs shared across study populations, the structure of response allowed for differentiation between different geographical areas and study design, revealing 3 clusters: (1) Uganda, Ecuador, and Brazil; (2) U-BIOPRED children and adults; (3) New Zealand, UK and MAAS. Spectral clustering identified differences between clusters, with 5 c-sIgEs groups solution characterizing Cluster 1. Finally, we observed almost parallel and constant shifts between severe and non-severe asthma subjects in each country.Patterns of c-sIgE responses are reflective of geographical location and study design. However, despite geographical differences in c-sIgE networks, we observed a remarkably consistent shift between networks of subjects with mild/moderate and severe asthma.CONCLUSIONSPatterns of c-sIgE responses are reflective of geographical location and study design. However, despite geographical differences in c-sIgE networks, we observed a remarkably consistent shift between networks of subjects with mild/moderate and severe asthma.