Dhouib et al. (DeepPlantAllergy: deep learning for explainable prediction of allergenicity in plant proteins. Brief Bioinform 2025;26:bbaf605.) developed DeepPlantAllergy, a deep learning model for predicting allergenicity in plant proteins, reporting area under the receiver operating characteristic curve (ROC-AUC) ≈ 97.7–97.8% on an independent test set. However, the dataset construction may lead to optimistic performance estimates. Specifically, non-allergen sequences sharing >20% identity with allergens were removed before the train/test split, which can reduce the presence of “hard negatives” (moderately similar non-allergens) in the test set and thereby weaken assessment under realistic screening conditions. Because practical allergen screening requires discrimination against large numbers of non-allergens that may share moderate sequence identity, we suggest re-evaluating the model using test sets that retain challenging negatives (with filtering performed against training allergens only) and reporting precision–recall metrics (area under the precision–recall curve) alongside ROC-AUC to better reflect performance under class imbalance.

I read with great interest the article by Yorulmaz et al

Background: Mosquito bite hypersensitivity results from abnormal immune responses to salivary proteins with potent immunomodulatory properties. In children, both type I (immunoglobulin E [IgE] mediated) and type IV (cell mediated) mechanisms contribute to the reaction spectrum, ranging from localized erythema to extensive inflammatory lesions. Although mosquito-specific IgE assays are commercially available, their diagnostic utility and correlation with clinical severity remain unclear. Objective: The objective was to describe the clinical and immunologic characteristics of children with mosquito hypersensitivity, evaluate the diagnostic relevance of mosquito-specific IgE, and explore associations with atopic background. Methods: A retrospective study was conducted on 206 children who underwent mosquito-specific IgE testing between June 2023 and June 2025. The mosquito allergy group included children with large local reactions, prolonged lesions, or systemic symptoms; the controls had typical reactions only. Clinical features, atopic comorbidities, total IgE levels, eosinophil percentages, and inhalant sensitizations were compared between the groups. Results: Atopic diseases were significantly more common in the mosquito allergy group versus the controls (87.0% versus 52.7%; p < 0.001). Mosquito-specific IgE positivity was low and similar across the groups (mosquito allergy group 6/77 [7.8%] vs controls 13/129 [10.1%]; p = 0.63), which showed no correlation with reaction type, duration, or severity. The total IgE levels were significantly higher in the mosquito allergy group (p = 0.045). The most frequent lesions were erythematous papules (88.3%) and immediate wheals (89.6%), whereas bullous lesions were rare (7.8%) and self-limited. Quality of life was mainly affected by avoidance of outdoor activities (16.9%) and psychological distress (15.6%). Conclusion: In pediatric mosquito hypersensitivity, clinical assessment outweighs mosquito-specific IgE testing for diagnosis. Despite the strong association with atopy, specific IgE positivity remains low, which supports the primacy of clinical history and lesion morphology in diagnostic decision-making.