Deep-AntiFP: Prediction of antifungal peptides using distanct multi-informative features incorporating with deep neural networks

World widely, Fungal infections have become a serious issue for human beings. Fungal infections normally happen once invading fungus appear on a specific area of the body and become difficult for the human immune system to control. Current antifungal based therapies and drugs are deemed unsatisfactory due to their harmful side effects. Due to the rapid increase of this toxic disease in the human body, accurate identification of AFPs has become a challenging task for the researchers. Measuring the effectiveness of AFPs on the human body, a reliable intelligent model is highly indispensable for the accurate identification of Antifungal Peptides. In this study, high discriminative numerical descriptors are extracted from peptide sequences using composite physiochemical properties, quasi sequence order, and reduce amino acid alphabet. Furthermore, a multi-information fusion approach is also utilized for compensating the weakness of individual feature spaces. Finally, several classification learning methods are applied to choose a suitable operational engine for our model. After evaluating the empirical results, deep neural networks with hyper-parameter optimal values achieved an accuracy of 94.23%, 91.02%, and 89.08% using the training, alternate and independent dataset, respectively. It was found that our proposed "DeepAntiFP" outperformed and reported the highest performance than the existing computational models. It is expected that the developed model may be played a useful role in research academia as well as proteomics and drug development. The source code and all datasets are publicly available at https://github.com/shahidawkum/Deep-AntiFP.