Assessing the Predictive Ability of Computational Epitope Prediction Methods on Fel d 1 Allergen

Purpose Though computational epitope prediction methods have been widely used, there have been only limited studies conducted in the context of allergies. Our research aims to benchmark publicly available epitope prediction tools, focusing on Fel d 1, whose allergenic IgE- and T-cell epitopes have been extensively studied. Methods Our study utilized an array of epitope prediction tools publicly accessible via the Immune Epitope Database (IEDB) and other resources. The tools were evaluated based on their ability to identify known linear IgE- and T-cell epitopes of Fel d 1. Results In general, B-cell epitope prediction methods demonstrated limited effectiveness. Most methods perform marginally better than random selection. ElliPro, a structure-based method, slightly outperformed the rest, suggesting that incorporating 3D structure information could enhance prediction accuracy. In terms of T-cell epitope prediction, ProPred was successful in identifying all known T-cell epitopes, whereas the IEDB approach missed two known epitopes and showed a high rate of over-prediction. Conclusions Our results show that current computational epitope prediction methods possess limitations in accurately identifying allergenic Fel d 1 epitopes. The study highlights the scope for future advancements in computational epitope prediction methodologies and the development of extensive epitope databases to optimize allergenic epitope prediction tools. Despite the evident limitations, these tools can still provide valuable preliminary insights into potential allergenic regions within proteins.