A Deep Learning Based Method to Predict Umami Peptides. Original paper:
Umami-MRNN: Deep learning-based prediction of umami peptide using RNN and MLP.
Lulu Qi, Jialuo Du, Yue Sun, Yongzhao Xiong, Xinyao Zhao, Daodong Pan, Yueru Zhi, Yali Dang, Xinchang Gao
Food Chemistry, 2023
| # | PEPTIDE NAME | PREDICTION RESULT |
|---|
umami-MRNN is a deep learning based method to predict umami peptides using combinations of features calculated from the peptide sequence. We extract not only the widely used statistical features including amino acid composition (AAC), dipeptide composition (DPC), dipeptide deviation from expected mean (DDE), and composition-transition-distribution (CTD), but we also include the peptide sequence itself encoded with overlapping property features (OPF) and one-hot codes. We use a merged model of The Multilayer Perceptron(MLP) and recursive neural network(RNN).
The Umami-MRNN Model of Umami Peptide Prediction