Abstract
Apples are highly susceptible to fungal infections, particularly by Alternaria species, which can lead to fruit deterioration and mycotoxin contamination during storage. This study aimed to evaluate the potential of untargeted liquid chromatography-high-resolution mass spectrometry (LC-HRMS) as a control-oriented strategy to detect Alternaria-infected apples under retail and long-term storage conditions. Healthy Red Delicious apples were artificially inoculated with three Alternaria tenuissima strains on the fruit surface or core and incubated at 25 degrees C or 4 degrees C. Extracts were analysed by UPLC-HRMS in both positive and negative electrospray ionisation modes, followed by multivariate chemometric analysis. Principal component analysis and partial least squares discriminant analysis consistently discriminated infected from non-infected apples, independent of strain, infection site, or incubation temperature. Feature selection based on variable importance values significantly improved model robustness and predictive performance. The metabolomic profiles also enabled discrimination according to Alternaria strain, infection site, storage temperature, and selected combinations of these factors. The results demonstrate that LC-HRMS-based untargeted metabolomics could provide a statistically robust framework for detecting Alternaria tenuissima infection in apples under the studied conditions.