Protein Secondary Structure Prediction Using Deep Convolutional Neural Fields

By Sheng Wang, Jian Peng, Jianzhu Ma, Jinbo Xu. 2015

544 citations recorded by the source corpus.

Protein secondary structure (SS) prediction is important for studying protein structure and function. When only the sequence (profile) information is used as input feature, currently the best predictors can obtain ~80% Q3 accuracy, which has not been improved in the past decade. Here we present DeepCNF (Deep Convolutional Neural Fields) for protein SS prediction. DeepCNF is a Deep Learning extension of Conditional Neural Fields (CNF), which is an integration of Conditional Random Fields (CRF) and shallow neural networks. DeepCNF can model not only complex sequence-structure relationship by a deep hierarchical architecture, but also interdependency between adjacent SS labels, so it is much more powerful than CNF. Experimental results show that DeepCNF can obtain ~84% Q3 accuracy, ~85% SOV score and ~72% Q8 accuracy, respectively, on the CASP and CAMEO test proteins, greatly outperforming currently popular predictors. As a general framework, DeepCNF can be used to predict other protein structure properties such as contact number, disorder regions and solvent accessibility.

Subfields: Computational Theory and Mathematics, Molecular Biology, Model Architectures, Computational Biology & Bioinformatics

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computational Theory and Mathematics

In Molecular Biology

In Model Architectures

In Computational Biology & Bioinformatics

Citations in the corpus

Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences (2019, methodology): “For calibration, current state-of-the-art on the same dataset is 84 % on the Q3 (3-class) task (Wang et al., 2016) or 70.5% on Q8 (8-class) task (Drori et al., 2018) using a technique considerably more sophisticated than linear projection.”
Modeling aspects of the language of life through transfer-learning protein sequences (2019, methodology): “Performance values for the same data set exist also for other recent methods such as Spider3 [62], RaptorX [63, 64] and JPred4 [65].”
Distance-based protein folding powered by deep learning (2018, background): “Sequential features include sequence profile and secondary structure as well as solvent accessibility predicted by RaptorX-Property (42).”
DeepSF: deep convolutional neural network for mapping protein sequences to folds (2017, methodology): “eep residual convolutional neural network was designed to further improve the accuracy of contact prediction [23]. Deep learning methods have also been applied to predict protein secondary structures [24, 25] and identify protein pairs that have the same fold [8, 14]. Here, we design a one-dimension”
Accurate De Novo Prediction of Protein Contact Map by Ultra-Deep Learning Model (2016, neutral): “com/convolutional-neural-networks/, and paper [44] for an example.”
Folding membrane proteins by deep transfer learning (2017, methodology): “The sequential features include protein sequence profile and secondary structure predicted by RaptorX-Property (Wang et al., 2016b, Wang et al., 2016a).”
RaptorX-Property: a web server for protein structure property prediction (2016, methodology): “To predict structure properties, this server employs a new machine learning model DeepCNF (Deep Convolutional Neural Fields) (13), which embraces the advantages of both conditional neural fields (CNF) (14) and deep convolutional neural networks (DCNN) (15).”
Spherical convolutions and their application in molecular modelling (2017, background): “…have demonstrated interesting examples in a broad selection of the natural sciences, ranging from physics (Aurisano et al., 2016; Mills et al., 2017) to biology (Wang et al., 2016; Min et al., 2017), in many cases showing convolutional neural networks to substantially outperform existing methods.”
ComplexContact: a web server for inter-protein contact prediction using deep learning (2018, background): “The sequential features include protein sequence profile, predicted 3-state secondary structure (31) and 3-state sol-W434 vent accessibility (32,33).”
Applications of Deep Learning and Reinforcement Learning to Biological Data (2017, unknown): “utilized deep CNN to predict 2-PS [98].”
RaptorX-Angle: real-value prediction of protein backbone dihedral angles through a hybrid method of clustering and deep learning (2018, methodology): “Inspired by the excellent performance of convolution neural network in predicting secondary structure [31] and order/disorder regions [32] and also the success of residual framework to do contact prediction [33], we adopt the ultra deep residual framework of convolutional neural network to do…”