Improving prediction of secondary structure, local backbone angles, and solvent accessible surface area of proteins by iterative deep learning

By Rhys Heffernan, Kuldip K. Paliwal, James G. Lyons, Abdollah Dehzangi, Alok Sharma, Jihua Wang, Abdul Sattar, Yuedong Yang, Yaoqi Zhou. 2015

346 citations recorded by the source corpus.

Direct prediction of protein structure from sequence is a challenging problem. An effective approach is to break it up into independent sub-problems. These sub-problems such as prediction of protein secondary structure can then be solved independently. In a previous study, we found that an iterative use of predicted secondary structure and backbone torsion angles can further improve secondary structure and torsion angle prediction. In this study, we expand the iterative features to include solvent accessible surface area and backbone angles and dihedrals based on Cα atoms. By using a deep learning neural network in three iterations, we achieved 82% accuracy for secondary structure prediction, 0.76 for the correlation coefficient between predicted and actual solvent accessible surface area, 19° and 30° for mean absolute errors of backbone φ and ψ angles, respectively and 8° and 32° for mean absolute errors of Cα-based θ and τ angles, respectively, for an independent test dataset of 1199 proteins. The accuracy of the method is slightly lower for 72 CASP 11 targets but much higher than those of model structures from current state-of-the-art techniques. This suggests the potentially beneficial use of these predicted properties for model assessment and ranking.

Subfields: Materials Chemistry, Molecular Biology

How researchers see this paper

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

In Materials Chemistry

1. Jinbo Xu
2. Sheng Wang
3. Chaopeng Shen
4. Kuldip K. Paliwal
5. Yaoqi Zhou
6. Yuedong Yang
7. Jack Hanson
8. Quan Le

In Molecular Biology

1. Jinbo Xu
2. Sheng Wang
3. Yaoqi Zhou
4. Kuldip K. Paliwal
5. Jack Hanson
6. Yuedong Yang
7. Yang Zhang
8. Jianlin Cheng

Citations in the corpus

• Opportunities and obstacles for deep learning in biology and medicine (2017, methodology): “Several groups [28,191,192] initiated the application of deep learning to protein secondary structure prediction but were unable to achieve significant improvement over the de facto standard method PSIPRED [193], which uses two shallow feedforward neural networks.”
• SPOT-Disorder2: Improved Protein Intrinsic Disorder Prediction by Ensembled Deep Learning (2019, neutral): “[40,41].”
• RaptorX-Property: a web server for protein structure property prediction (2016, methodology): “In the future, we may further improve prediction accuracy by extending our DeepCNF into a multi-task-enabled deep learning model (31).”
• A Transdisciplinary Review of Deep Learning Research and Its Relevance for Water Resources Scientists (2017, methodology): “Stacked auto-encoders and DBNs were trained to predict protein structures directly from genome sequence with much better accuracy [Lyons et al., 2014; Heffernan et al., 2015; Spencer et al., 2015].”
• Extant fold-switching proteins are widespread (2018, background): “On a related note, our predictions suggest that the ∼80% plateau in secondary structure prediction accuracy (35) might be due, in part, to fold-switching regions in proteins.”
• Deep learning in bioinformatics (2016, methodology): “[85] applied SAE to protein amino acid sequences to solve prediction problems for secondary structure, torsion angle and accessible surface area.”
• Deep Learning and Its Applications in Biomedicine (2018, positive): “Heffernan et al. also employed SAEs to predict secondary structure, local backbone angles, and solvent-accessible surface area (ASA) of proteins from amino acid sequences [170]; they achieved an accuracy of 82% for secondary structure prediction.”
• RaptorX-Angle: real-value prediction of protein backbone dihedral angles through a hybrid method of clustering and deep learning (2018, methodology): “Zhou et al. had developed SPIDER2 [27] to predict real-valued angles and then separately SPIDER2-Delta [35] to predict error of those predicted structural properties.”
• Sixty-five years of the long march in protein secondary structure prediction: the final stretch? (2016, methodology): “Can we make the last stretch of this long-standing challenge? Recent accuracy improvements [91–93, 95] have resulted from increasingly larger sequence [4] and structural databases [5, 145], more sophisticated deep learning neural networks [57, 58], and the use of structural template information in w”
• Applications of Deep Learning and Reinforcement Learning to Biological Data (2017, methodology): “proposed an iterative DNN scheme to solve these sub-problems for 2-PS [87].”