DeepTP: A Deep Learning Model for Thermophilic Protein Prediction

• Published in: International journal of molecular sciences 2023-01, Vol.24 (3), p.2217
• Main Authors: Zhao, Jianjun, Yan, Wenying, Yang, Yang
• Format: Article
• Language: English
• Subjects: Algorithms; Amino acid sequence; Amino acids; bidirectional long short-term memory network; convolutional neural network; Datasets; Deep Learning; Feature selection; Heat resistance; Hydrogen bonds; Industrial production; Information processing; Long short-term memory; Machine Learning; Methods; multiple-channel feature fusion; Neural networks; Neural Networks, Computer; Performance evaluation; Pharmaceuticals; Prediction models; Protein engineering; Proteins; Proteins - metabolism; self-attention; Support vector machines; Test sets; thermophilic proteins
• ISSN: 1422-0067; 1661-6596; 1422-0067
• DOI: 10.3390/ijms24032217

Thermophilic proteins have important value in the fields of biopharmaceuticals and enzyme engineering. Most existing thermophilic protein prediction models are based on traditional machine learning algorithms and do not fully utilize protein sequence information. To solve this problem, a deep learning model based on self-attention and multiple-channel feature fusion was proposed to predict thermophilic proteins, called DeepTP. First, a large new dataset consisting of 20,842 proteins was constructed. Second, a convolutional neural network and bidirectional long short-term memory network were used to extract the hidden features in protein sequences. Different weights were then assigned to features through self-attention, and finally, biological features were integrated to build a prediction model. In a performance comparison with existing methods, DeepTP had better performance and scalability in an independent balanced test set and validation set, with AUC values of 0.944 and 0.801, respectively. In the unbalanced test set, DeepTP had an average precision (AP) of 0.536. The tool is freely available.