CBEIL17:IL-17诱导肽预测的深度学习方法OA

Interleukin-17(IL-17)is an important glycoprotein whose sustained production is closely related to the pathological mechanisms of various autoimmune and inflammatory diseases.Meanwhile,it also plays an indispensable role in the defense mechanism of the body against extracellular pathogen infections.Therefore,accurate identification of IL-17-inducing peptides is of great significance for the study of IL-17.Since the biological experimental identification of IL-17-inducing peptides is time-consuming and costly,this study proposes a prediction method based on deep learning and an image feature enhancement strat-egy to improve the identification efficiency of IL-17-inducing peptides.First,multi-view feature encoding is performed on IL-17-inducing peptide sequences,including conventional feature descriptors,pretrained embedding features from protein lan-guage models,cellular automata image encoding features,and local structural features.Subsequently,the multi-view features are processed based on the image feature enhancement strategy,and the enhanced features are combined with the original fea-tures.Finally,the combined features are input into a convolutional neural network and a bidirectional temporal convolutional net-work to further extract deep features and construct the model.Experimental results show that this method achieves an AUC of 85.7%on the independent test dataset,demonstrating good predictive performance and providing a new tool for the identification of IL-17-inducing peptides.