Text this: The impact of biologically relevant negative samples on machine learning-based B-cell epitope prediction for Influenza A