Protein-protein binding affinity prediction based on an SVR ensemble

Xueling Li; Min Zhu; Xiaolai Li; Hong Qiang Wang; Shulin Wang

doi:10.1007/978-3-642-31588-6_19

Protein-protein binding affinity prediction based on an SVR ensemble

Xueling Li, Min Zhu, Xiaolai Li, Hong Qiang Wang, Shulin Wang

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

12 Scopus citations

Abstract

Accurately predicting generic protein-protein binding affinities (PPBA) is essential to analyze the outputs of protein docking and may help infer real status of cellular protein-protein interaction sub-networks. However, accurate PPBA prediction is still extremely challenging. Machine learning methods are promising to address this problem. We propose a two-layer support vector regression (TLSVR) model to implicitly capture binding contributions that are hard to explicitly model. The TLSVR circumvents both the descriptor compatibility problem and the need for problematic modeling assumptions. Input features for TLSVR in first layer are scores of 2209 interacting atom pairs within each distance bin. The base SVRs are combined by the second layer to infer the final affinities. Leave-one-out validation on our heterogeneous data shows that the TLSVR method obtains a very good result of R=0.80 and SD=1.32 with real affinities. Comparison experiment further demonstrates that TLSVR is superior to the previous state-of-art methods in predicting generic PPBA.

Original language	English (US)
Title of host publication	Intelligent Computing Technology - 8th International Conference, ICIC 2012, Proceedings
Pages	145-151
Number of pages	7
DOIs	https://doi.org/10.1007/978-3-642-31588-6_19
State	Published - 2012
Externally published	Yes
Event	8th International Conference on Intelligent Computing Technology, ICIC 2012 - Huangshan, China Duration: Jul 25 2012 → Jul 29 2012

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	7389 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Other

Other	8th International Conference on Intelligent Computing Technology, ICIC 2012
Country/Territory	China
City	Huangshan
Period	7/25/12 → 7/29/12

Keywords

Protein-protein interaction affinity
machine learning
potential of mean force
two-layer support vector machine

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-642-31588-6_19

Cite this

Li, X., Zhu, M., Li, X., Wang, H. Q., & Wang, S. (2012). Protein-protein binding affinity prediction based on an SVR ensemble. In Intelligent Computing Technology - 8th International Conference, ICIC 2012, Proceedings (pp. 145-151). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7389 LNCS). https://doi.org/10.1007/978-3-642-31588-6_19

Protein-protein binding affinity prediction based on an SVR ensemble. / Li, Xueling; Zhu, Min; Li, Xiaolai et al.
Intelligent Computing Technology - 8th International Conference, ICIC 2012, Proceedings. 2012. p. 145-151 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7389 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Li, X, Zhu, M, Li, X, Wang, HQ & Wang, S 2012, Protein-protein binding affinity prediction based on an SVR ensemble. in Intelligent Computing Technology - 8th International Conference, ICIC 2012, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 7389 LNCS, pp. 145-151, 8th International Conference on Intelligent Computing Technology, ICIC 2012, Huangshan, China, 7/25/12. https://doi.org/10.1007/978-3-642-31588-6_19

Li X, Zhu M, Li X, Wang HQ, Wang S. Protein-protein binding affinity prediction based on an SVR ensemble. In Intelligent Computing Technology - 8th International Conference, ICIC 2012, Proceedings. 2012. p. 145-151. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-642-31588-6_19

@inproceedings{75e00ed3bd994ebda8b0b62621a7d1d1,

title = "Protein-protein binding affinity prediction based on an SVR ensemble",

abstract = "Accurately predicting generic protein-protein binding affinities (PPBA) is essential to analyze the outputs of protein docking and may help infer real status of cellular protein-protein interaction sub-networks. However, accurate PPBA prediction is still extremely challenging. Machine learning methods are promising to address this problem. We propose a two-layer support vector regression (TLSVR) model to implicitly capture binding contributions that are hard to explicitly model. The TLSVR circumvents both the descriptor compatibility problem and the need for problematic modeling assumptions. Input features for TLSVR in first layer are scores of 2209 interacting atom pairs within each distance bin. The base SVRs are combined by the second layer to infer the final affinities. Leave-one-out validation on our heterogeneous data shows that the TLSVR method obtains a very good result of R=0.80 and SD=1.32 with real affinities. Comparison experiment further demonstrates that TLSVR is superior to the previous state-of-art methods in predicting generic PPBA.",

keywords = "Protein-protein interaction affinity, machine learning, potential of mean force, two-layer support vector machine",

author = "Xueling Li and Min Zhu and Xiaolai Li and Wang, {Hong Qiang} and Shulin Wang",

year = "2012",

doi = "10.1007/978-3-642-31588-6_19",

language = "English (US)",

isbn = "9783642315879",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

pages = "145--151",

booktitle = "Intelligent Computing Technology - 8th International Conference, ICIC 2012, Proceedings",

note = "8th International Conference on Intelligent Computing Technology, ICIC 2012 ; Conference date: 25-07-2012 Through 29-07-2012",

}

TY - GEN

T1 - Protein-protein binding affinity prediction based on an SVR ensemble

AU - Li, Xueling

AU - Zhu, Min

AU - Li, Xiaolai

AU - Wang, Hong Qiang

AU - Wang, Shulin

PY - 2012

Y1 - 2012

N2 - Accurately predicting generic protein-protein binding affinities (PPBA) is essential to analyze the outputs of protein docking and may help infer real status of cellular protein-protein interaction sub-networks. However, accurate PPBA prediction is still extremely challenging. Machine learning methods are promising to address this problem. We propose a two-layer support vector regression (TLSVR) model to implicitly capture binding contributions that are hard to explicitly model. The TLSVR circumvents both the descriptor compatibility problem and the need for problematic modeling assumptions. Input features for TLSVR in first layer are scores of 2209 interacting atom pairs within each distance bin. The base SVRs are combined by the second layer to infer the final affinities. Leave-one-out validation on our heterogeneous data shows that the TLSVR method obtains a very good result of R=0.80 and SD=1.32 with real affinities. Comparison experiment further demonstrates that TLSVR is superior to the previous state-of-art methods in predicting generic PPBA.

AB - Accurately predicting generic protein-protein binding affinities (PPBA) is essential to analyze the outputs of protein docking and may help infer real status of cellular protein-protein interaction sub-networks. However, accurate PPBA prediction is still extremely challenging. Machine learning methods are promising to address this problem. We propose a two-layer support vector regression (TLSVR) model to implicitly capture binding contributions that are hard to explicitly model. The TLSVR circumvents both the descriptor compatibility problem and the need for problematic modeling assumptions. Input features for TLSVR in first layer are scores of 2209 interacting atom pairs within each distance bin. The base SVRs are combined by the second layer to infer the final affinities. Leave-one-out validation on our heterogeneous data shows that the TLSVR method obtains a very good result of R=0.80 and SD=1.32 with real affinities. Comparison experiment further demonstrates that TLSVR is superior to the previous state-of-art methods in predicting generic PPBA.

KW - Protein-protein interaction affinity

KW - machine learning

KW - potential of mean force

KW - two-layer support vector machine

UR - http://www.scopus.com/inward/record.url?scp=84865305788&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84865305788&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-31588-6_19

DO - 10.1007/978-3-642-31588-6_19

M3 - Conference contribution

AN - SCOPUS:84865305788

SN - 9783642315879

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 145

EP - 151

BT - Intelligent Computing Technology - 8th International Conference, ICIC 2012, Proceedings

T2 - 8th International Conference on Intelligent Computing Technology, ICIC 2012

Y2 - 25 July 2012 through 29 July 2012

ER -

Protein-protein binding affinity prediction based on an SVR ensemble

Abstract

Publication series

Other

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this