Abstract:The prediction of secondary structure is an important topic in the field of bioinformatics, even if the methods have matured, and development of the algorithms is a far less active area than a decade ago. Accurate prediction is very useful to biologists in its own right, but it is worth pointing out that it is also an essential component of tertiary structure prediction, which in contrast is far from solved and continues to be a highly active area of research. In addition, sequence comparison methods have more recently incorporated local structure tracks. The extra information utilized by the new methods has led to considerable improvements in fold recognition and alignment accuracy. In this paper, a novel method for protein secondary structure prediction is presented. Using evolutionary information contained in amino acid's physicochemical properties, position-specific scoring matrix generated by PSI-BLAST and HMMER3 profiles as input to hybrid back propagation system, secondary structure can be predicted at significantly increased accuracy. Based on knowledge discovery theory based on inner cognitive mechanism (KDTICM) theory, we have constructed a compound pyramid model approach, which is composed of four layers of the intelligent interface and integrated in several ways, such as hybrid back propagation method (HBP), modified knowledge discovery in databases (KDD*), hybrid SVM method (HSVM) and so on. Experiments on three standard datasets (RS126, CB513 and CASP8) show that CPM is capable of producing the higher Q (3) and SOV scores than that achieved by existing widely used schemes such as PSIPRED, PHD, Predator, as well as previously developed prediction methods. On the RS126 and CB513 datasets, it achieves a Q (3) and SOV99 score are considerably higher than the best reported scores, respectively. It is also tested on target proteins of critical assessment of protein structure prediction experiment (CASP8) and achieves better results than the traditional methods, including the popular PSIPRED method over overall prediction accuracy. Available: http://www.kdd.ustb.edu.cn/protein_Web.

A Deep Aggregated Model for Protein Secondary Structure Prediction

Sequence Translating Model Using Deep Neural Block Cascade Network: Taking Protein Secondary Structure Prediction As an Example

A Deep Convolutional Neural Network to Improve the Prediction of Protein Secondary Structure

Protein secondary structure prediction by using deep learning method.

DeepACLSTM: deep asymmetric convolutional long short-term memory neural models for protein secondary structure prediction

Localnet: A Simple Recurrent Neural Network Model for Protein Secondary Structure Prediction Using Local Amino Acid Sequences Only

Protein secondary structure prediction using deep convolutional neural fields

Deep Ensemble Learning with Atrous Spatial Pyramid Networks for Protein Secondary Structure Prediction

Improved Protein Secondary Structure Prediction Using Bidirectional Long Short-Term Memory Neural Network and Bootstrap Aggregating

Prediction of 8-State Protein Secondary Structures by a Novel Deep Learning Architecture

PS8-Net: A Deep Convolutional Neural Network to Predict the Eight-State Protein Secondary Structure

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

Protein Secondary Structure Prediction Using Deep Multi-scale Convolutional Neural Networks and Next-Step Conditioning

Predicting protein second structure using a novel hybrid method

Deep Supervised and Convolutional Generative Stochastic Network for Protein Secondary Structure Prediction

Combining GCN and Bi-LSTM for Protein Secondary Structure Prediction

HYBP_PSSP: a Hybrid Back Propagation Method for Predicting Protein Secondary Structure.

Protein Secondary Structure Prediction Based on Cascade Neural Networks

A Highly Targeted Autoencoder-Driven Deep Multi-Task Learning Framework for Accurate Protein Secondary Structure Prediction

Predicting Protein Secondary Structure Based on Compound Pyramid Model.

Comprehensive Study on Enhancing Low-Quality Position-Specific Scoring Matrix with Deep Learning for Accurate Protein Structure Property Prediction: Using Bagging Multiple Sequence Alignment Learning