Spotlight: Assembly of protein complexes by integrating graph clustering methods

Chia Hao Chin, Shu Hwa Chen, Chun Yu Chen, Chao A. Hsiung, Chin Wen Ho, Ming Tat Ko, Chung Yen Lin

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

As is generally assumed, clusters in protein-protein interaction (PPI) networks perform specific, crucial functions in biological systems. Various network community detection methods have been developed to exploit PPI networks in order to identify protein complexes and functional modules. Due to the potential role of various regulatory modes in biological networks, a single method may just apply a single graph property and neglect communities highlighted by other network properties. This work presents a novel integration method to capture protein modules/protein complexes by multiple network features detected by different algorithms. The integration method is further implemented in a web-based platform with a highly effective interactive network analyzer. Conventionally adopted methods with different perspectives on network community detection (e.g., CPM, FastGreedy, HUNTER, MCL, LE, SpinGlass, and WalkTrap) are also executed simultaneously. Analytical results indicate that the proposed method performs better than the conventional ones. The proposed approach can capture the transcription and RNA splicing machineries from the yeast protein network. Meanwhile, proteins that are highly associated with each other, yet not described in both machineries are also identified. In sum, a protein that is closely connected to components of a known module or a complex in the network view implies the functional association among them. Importantly, our method can detect these unique network features, thus facilitating efforts to discover unknown components of functional modules/protein complexes. Availability: Spotlight is freely accessible at http://hub.iis.sinica.edu.tw/spotlight. Video clips for a quick view of usage are available in the website online help page.

Original languageEnglish
Pages (from-to)42-51
Number of pages10
JournalGene
Volume518
Issue number1
DOIs
Publication statusPublished - Apr 10 2013
Externally publishedYes

Fingerprint

Cluster Analysis
Proteins
Protein Interaction Maps
RNA Splicing
Fungal Proteins
Surgical Instruments

Keywords

  • Algorithm
  • Network biology
  • Protein complex
  • Topology

ASJC Scopus subject areas

  • Genetics

Cite this

Chin, C. H., Chen, S. H., Chen, C. Y., Hsiung, C. A., Ho, C. W., Ko, M. T., & Lin, C. Y. (2013). Spotlight: Assembly of protein complexes by integrating graph clustering methods. Gene, 518(1), 42-51. https://doi.org/10.1016/j.gene.2012.11.087

Spotlight : Assembly of protein complexes by integrating graph clustering methods. / Chin, Chia Hao; Chen, Shu Hwa; Chen, Chun Yu; Hsiung, Chao A.; Ho, Chin Wen; Ko, Ming Tat; Lin, Chung Yen.

In: Gene, Vol. 518, No. 1, 10.04.2013, p. 42-51.

Research output: Contribution to journalArticle

Chin, CH, Chen, SH, Chen, CY, Hsiung, CA, Ho, CW, Ko, MT & Lin, CY 2013, 'Spotlight: Assembly of protein complexes by integrating graph clustering methods', Gene, vol. 518, no. 1, pp. 42-51. https://doi.org/10.1016/j.gene.2012.11.087
Chin, Chia Hao ; Chen, Shu Hwa ; Chen, Chun Yu ; Hsiung, Chao A. ; Ho, Chin Wen ; Ko, Ming Tat ; Lin, Chung Yen. / Spotlight : Assembly of protein complexes by integrating graph clustering methods. In: Gene. 2013 ; Vol. 518, No. 1. pp. 42-51.
@article{d8ef6abcf5484008852030d80d4e79bb,
title = "Spotlight: Assembly of protein complexes by integrating graph clustering methods",
abstract = "As is generally assumed, clusters in protein-protein interaction (PPI) networks perform specific, crucial functions in biological systems. Various network community detection methods have been developed to exploit PPI networks in order to identify protein complexes and functional modules. Due to the potential role of various regulatory modes in biological networks, a single method may just apply a single graph property and neglect communities highlighted by other network properties. This work presents a novel integration method to capture protein modules/protein complexes by multiple network features detected by different algorithms. The integration method is further implemented in a web-based platform with a highly effective interactive network analyzer. Conventionally adopted methods with different perspectives on network community detection (e.g., CPM, FastGreedy, HUNTER, MCL, LE, SpinGlass, and WalkTrap) are also executed simultaneously. Analytical results indicate that the proposed method performs better than the conventional ones. The proposed approach can capture the transcription and RNA splicing machineries from the yeast protein network. Meanwhile, proteins that are highly associated with each other, yet not described in both machineries are also identified. In sum, a protein that is closely connected to components of a known module or a complex in the network view implies the functional association among them. Importantly, our method can detect these unique network features, thus facilitating efforts to discover unknown components of functional modules/protein complexes. Availability: Spotlight is freely accessible at http://hub.iis.sinica.edu.tw/spotlight. Video clips for a quick view of usage are available in the website online help page.",
keywords = "Algorithm, Network biology, Protein complex, Topology",
author = "Chin, {Chia Hao} and Chen, {Shu Hwa} and Chen, {Chun Yu} and Hsiung, {Chao A.} and Ho, {Chin Wen} and Ko, {Ming Tat} and Lin, {Chung Yen}",
year = "2013",
month = "4",
day = "10",
doi = "10.1016/j.gene.2012.11.087",
language = "English",
volume = "518",
pages = "42--51",
journal = "Gene",
issn = "0378-1119",
publisher = "Elsevier",
number = "1",

}

TY - JOUR

T1 - Spotlight

T2 - Assembly of protein complexes by integrating graph clustering methods

AU - Chin, Chia Hao

AU - Chen, Shu Hwa

AU - Chen, Chun Yu

AU - Hsiung, Chao A.

AU - Ho, Chin Wen

AU - Ko, Ming Tat

AU - Lin, Chung Yen

PY - 2013/4/10

Y1 - 2013/4/10

N2 - As is generally assumed, clusters in protein-protein interaction (PPI) networks perform specific, crucial functions in biological systems. Various network community detection methods have been developed to exploit PPI networks in order to identify protein complexes and functional modules. Due to the potential role of various regulatory modes in biological networks, a single method may just apply a single graph property and neglect communities highlighted by other network properties. This work presents a novel integration method to capture protein modules/protein complexes by multiple network features detected by different algorithms. The integration method is further implemented in a web-based platform with a highly effective interactive network analyzer. Conventionally adopted methods with different perspectives on network community detection (e.g., CPM, FastGreedy, HUNTER, MCL, LE, SpinGlass, and WalkTrap) are also executed simultaneously. Analytical results indicate that the proposed method performs better than the conventional ones. The proposed approach can capture the transcription and RNA splicing machineries from the yeast protein network. Meanwhile, proteins that are highly associated with each other, yet not described in both machineries are also identified. In sum, a protein that is closely connected to components of a known module or a complex in the network view implies the functional association among them. Importantly, our method can detect these unique network features, thus facilitating efforts to discover unknown components of functional modules/protein complexes. Availability: Spotlight is freely accessible at http://hub.iis.sinica.edu.tw/spotlight. Video clips for a quick view of usage are available in the website online help page.

AB - As is generally assumed, clusters in protein-protein interaction (PPI) networks perform specific, crucial functions in biological systems. Various network community detection methods have been developed to exploit PPI networks in order to identify protein complexes and functional modules. Due to the potential role of various regulatory modes in biological networks, a single method may just apply a single graph property and neglect communities highlighted by other network properties. This work presents a novel integration method to capture protein modules/protein complexes by multiple network features detected by different algorithms. The integration method is further implemented in a web-based platform with a highly effective interactive network analyzer. Conventionally adopted methods with different perspectives on network community detection (e.g., CPM, FastGreedy, HUNTER, MCL, LE, SpinGlass, and WalkTrap) are also executed simultaneously. Analytical results indicate that the proposed method performs better than the conventional ones. The proposed approach can capture the transcription and RNA splicing machineries from the yeast protein network. Meanwhile, proteins that are highly associated with each other, yet not described in both machineries are also identified. In sum, a protein that is closely connected to components of a known module or a complex in the network view implies the functional association among them. Importantly, our method can detect these unique network features, thus facilitating efforts to discover unknown components of functional modules/protein complexes. Availability: Spotlight is freely accessible at http://hub.iis.sinica.edu.tw/spotlight. Video clips for a quick view of usage are available in the website online help page.

KW - Algorithm

KW - Network biology

KW - Protein complex

KW - Topology

UR - http://www.scopus.com/inward/record.url?scp=84875379772&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84875379772&partnerID=8YFLogxK

U2 - 10.1016/j.gene.2012.11.087

DO - 10.1016/j.gene.2012.11.087

M3 - Article

C2 - 23274651

AN - SCOPUS:84875379772

VL - 518

SP - 42

EP - 51

JO - Gene

JF - Gene

SN - 0378-1119

IS - 1

ER -