The article introduces Vaxijen, a server for predicting protective antigens, tumor antigens, and subunit vaccines. The authors, Irinia A. Doytchinova and Darren R. Flower, propose an alignment-free method based on auto cross covariance (ACC) transformation of protein sequences into uniform vectors of principal amino acid properties. This approach overcomes the limitations of sequence alignment methods, which can fail to identify antigens with subtle or novel sequences. The models were trained and validated using bacterial, viral, and tumor protein datasets, achieving prediction accuracies of 70% to 89%. The Vaxijen server is freely available and can be used independently or in conjunction with other bioinformatics tools for reverse vaccinology. The server can predict the protective antigen status of protein sequences based on their physicochemical properties, making it a valuable resource for vaccine development.The article introduces Vaxijen, a server for predicting protective antigens, tumor antigens, and subunit vaccines. The authors, Irinia A. Doytchinova and Darren R. Flower, propose an alignment-free method based on auto cross covariance (ACC) transformation of protein sequences into uniform vectors of principal amino acid properties. This approach overcomes the limitations of sequence alignment methods, which can fail to identify antigens with subtle or novel sequences. The models were trained and validated using bacterial, viral, and tumor protein datasets, achieving prediction accuracies of 70% to 89%. The Vaxijen server is freely available and can be used independently or in conjunction with other bioinformatics tools for reverse vaccinology. The server can predict the protective antigen status of protein sequences based on their physicochemical properties, making it a valuable resource for vaccine development.