Abstract:Secondary structure prediction is a key step in understanding protein function and biological properties and is highly important in the fields of new drug development, disease treatment, bioengineering, etc. Accurately predicting the secondary structure of proteins helps to reveal how proteins are folded and how they function in cells. The application of deep learning models in protein structure prediction is particularly important because of their ability to process complex sequence information and extract meaningful patterns and features, thus significantly improving the accuracy and efficiency of prediction. In this study, a combined model integrating an improved temporal convolutional network (TCN), bidirectional long short-term memory (BiLSTM), and a multi-head attention (MHA) mechanism is proposed to enhance the accuracy of protein prediction in both eight-state and three-state structures. One-hot encoding features and word vector representations of physicochemical properties are incorporated. A significant emphasis is placed on knowledge distillation techniques utilizing the ProtT5 pretrained model, leading to performance improvements. The improved TCN, achieved through multiscale fusion and bidirectional operations, allows for better extraction of amino acid sequence features than traditional TCN models. The model demonstrated excellent prediction performance on multiple datasets. For the TS115, CB513 and PDB (2018-2020) datasets, the prediction accuracy of the eight-state structure of the six datasets in this paper reached 88.2%, 84.9%, and 95.3%, respectively, and the prediction accuracy of the three-state structure reached 91.3%, 90.3%, and 96.8%, respectively. This study not only improves the accuracy of protein secondary structure prediction but also provides an important tool for understanding protein structure and function, which is particularly applicable to resource-constrained contexts and provides a valuable tool for understanding protein structure and function.

Protein tertiary structure prediction using hidden Markov model based on lattice

On use of tertiary structure characters in hidden Markov models for protein fold prediction

Protein Folding Prediction in a Cubic Lattice in Hydrophobic-Polar Model

Advancing the Accuracy of Protein Fold Recognition by Utilizing Profiles from Hidden Markov Models.

Enhanced Hybrid Search Algorithm for Protein Structure Prediction Using the 3D-HP Lattice Model

Protein Fold Recognition based on Multi-view Modeling.

Hybrid Model Of Neural Network And Hidden Markov Model For Protein Secondary Structure Prediction

Protein Structure Prediction in the 3D HP Model Using Deep Reinforcement Learning

MLDH-Fold: Protein Fold Recognition Based on Multi-View Low-Rank Modeling

Hidden Markov Model for Protein Structural Class Prediction Based on MATLAB

Prediction of protein secondary structure by the hidden Markov model

Accurate Protein Structure Prediction by Embeddings and Deep Learning Representations

Predicting Protein Structural Class Based on Hidden Markov Models

Recent Progress of Protein Tertiary Structure Prediction

A Novel Protein Secondary Structure Prediction System Based on Compound Pyramid Model

CMB-2019-0193-ver9-Hanson_1P 1..20

MCP: a Multi-Component learning machine to Predict protein secondary structure

Protein secondary structure and remote homology detection

Algorithm for Predicting Protein Secondary Structure

HMM in Predicting Protein Secondary Structure

Prediction of protein secondary structure by the improved TCN-BiLSTM-MHA model with knowledge distillation