A dynamic Bayesian framwork to learn temporal gene interactions using external knowledge

Umut Aǧyüz, Şenol Işçi, Cengizhan Öztürk, Ahmet Ademoǧlu, Hasan H. Otu

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

One of the main problems in systems biology is learning gene interaction networks from experimental data. This turns out to be a challenging task as the experimental data is sparse and noisy, and network learning algorithms are computationally intense. Bayesian Networks (BN) have become a popular choice for learning such networks as BNs avoid overfitting and are robust to noise. In this paper we build up on our established framework, Bayesian Network Prior, where we incorporate existing biological knowledge in learning gene interaction networks. However, biological phenomena are time-dependent and there is need to extend the static structure of learning approaches to a temporal level. Here, we present a Dynamic BN framework, which learns interaction networks between different time points in time-series data. Both intra and inter networks are learnt and compared to standard DBN learning algorithms. Our results based on synthetic and simulated gene expression data suggest that the proposed method outperforms existing approaches in identifying the underlying network structure. The proposed framework is robust to errors in the incorporated knowledge and can combine various experimental data types together with existing knowledge when learning networks.

Original languageEnglish (US)
Title of host publication2013 8th International Symposium on Health Informatics and Bioinformatics, HIBIT 2013
PublisherIEEE Computer Society
ISBN (Print)9781479907014
DOIs
StatePublished - Jan 1 2013
Event2013 8th International Symposium on Health Informatics and Bioinformatics, HIBIT 2013 - Ankara, Turkey
Duration: Sep 25 2013Sep 27 2013

Publication series

Name2013 8th International Symposium on Health Informatics and Bioinformatics, HIBIT 2013

Other

Other2013 8th International Symposium on Health Informatics and Bioinformatics, HIBIT 2013
CountryTurkey
CityAnkara
Period9/25/139/27/13

Fingerprint

Learning
Genes
Gene Regulatory Networks
Synthetic Genes
Biological Phenomena
Systems Biology
Noise
Gene Expression

Keywords

  • Dynamic Bayesian Networks
  • External biological knowledge
  • Gene interaction networks
  • Microarray
  • Time-series data

ASJC Scopus subject areas

  • Health Informatics

Cite this

Aǧyüz, U., Işçi, Ş., Öztürk, C., Ademoǧlu, A., & Otu, H. H. (2013). A dynamic Bayesian framwork to learn temporal gene interactions using external knowledge. In 2013 8th International Symposium on Health Informatics and Bioinformatics, HIBIT 2013 [6661680] (2013 8th International Symposium on Health Informatics and Bioinformatics, HIBIT 2013). IEEE Computer Society. https://doi.org/10.1109/HIBIT.2013.6661680

A dynamic Bayesian framwork to learn temporal gene interactions using external knowledge. / Aǧyüz, Umut; Işçi, Şenol; Öztürk, Cengizhan; Ademoǧlu, Ahmet; Otu, Hasan H.

2013 8th International Symposium on Health Informatics and Bioinformatics, HIBIT 2013. IEEE Computer Society, 2013. 6661680 (2013 8th International Symposium on Health Informatics and Bioinformatics, HIBIT 2013).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Aǧyüz, U, Işçi, Ş, Öztürk, C, Ademoǧlu, A & Otu, HH 2013, A dynamic Bayesian framwork to learn temporal gene interactions using external knowledge. in 2013 8th International Symposium on Health Informatics and Bioinformatics, HIBIT 2013., 6661680, 2013 8th International Symposium on Health Informatics and Bioinformatics, HIBIT 2013, IEEE Computer Society, 2013 8th International Symposium on Health Informatics and Bioinformatics, HIBIT 2013, Ankara, Turkey, 9/25/13. https://doi.org/10.1109/HIBIT.2013.6661680
Aǧyüz U, Işçi Ş, Öztürk C, Ademoǧlu A, Otu HH. A dynamic Bayesian framwork to learn temporal gene interactions using external knowledge. In 2013 8th International Symposium on Health Informatics and Bioinformatics, HIBIT 2013. IEEE Computer Society. 2013. 6661680. (2013 8th International Symposium on Health Informatics and Bioinformatics, HIBIT 2013). https://doi.org/10.1109/HIBIT.2013.6661680
Aǧyüz, Umut ; Işçi, Şenol ; Öztürk, Cengizhan ; Ademoǧlu, Ahmet ; Otu, Hasan H. / A dynamic Bayesian framwork to learn temporal gene interactions using external knowledge. 2013 8th International Symposium on Health Informatics and Bioinformatics, HIBIT 2013. IEEE Computer Society, 2013. (2013 8th International Symposium on Health Informatics and Bioinformatics, HIBIT 2013).
@inproceedings{f943557f4fb24d95a0d25b71a6893902,
title = "A dynamic Bayesian framwork to learn temporal gene interactions using external knowledge",
abstract = "One of the main problems in systems biology is learning gene interaction networks from experimental data. This turns out to be a challenging task as the experimental data is sparse and noisy, and network learning algorithms are computationally intense. Bayesian Networks (BN) have become a popular choice for learning such networks as BNs avoid overfitting and are robust to noise. In this paper we build up on our established framework, Bayesian Network Prior, where we incorporate existing biological knowledge in learning gene interaction networks. However, biological phenomena are time-dependent and there is need to extend the static structure of learning approaches to a temporal level. Here, we present a Dynamic BN framework, which learns interaction networks between different time points in time-series data. Both intra and inter networks are learnt and compared to standard DBN learning algorithms. Our results based on synthetic and simulated gene expression data suggest that the proposed method outperforms existing approaches in identifying the underlying network structure. The proposed framework is robust to errors in the incorporated knowledge and can combine various experimental data types together with existing knowledge when learning networks.",
keywords = "Dynamic Bayesian Networks, External biological knowledge, Gene interaction networks, Microarray, Time-series data",
author = "Umut Aǧy{\"u}z and Şenol Iş{\cc}i and Cengizhan {\"O}zt{\"u}rk and Ahmet Ademoǧlu and Otu, {Hasan H.}",
year = "2013",
month = "1",
day = "1",
doi = "10.1109/HIBIT.2013.6661680",
language = "English (US)",
isbn = "9781479907014",
series = "2013 8th International Symposium on Health Informatics and Bioinformatics, HIBIT 2013",
publisher = "IEEE Computer Society",
booktitle = "2013 8th International Symposium on Health Informatics and Bioinformatics, HIBIT 2013",

}

TY - GEN

T1 - A dynamic Bayesian framwork to learn temporal gene interactions using external knowledge

AU - Aǧyüz, Umut

AU - Işçi, Şenol

AU - Öztürk, Cengizhan

AU - Ademoǧlu, Ahmet

AU - Otu, Hasan H.

PY - 2013/1/1

Y1 - 2013/1/1

N2 - One of the main problems in systems biology is learning gene interaction networks from experimental data. This turns out to be a challenging task as the experimental data is sparse and noisy, and network learning algorithms are computationally intense. Bayesian Networks (BN) have become a popular choice for learning such networks as BNs avoid overfitting and are robust to noise. In this paper we build up on our established framework, Bayesian Network Prior, where we incorporate existing biological knowledge in learning gene interaction networks. However, biological phenomena are time-dependent and there is need to extend the static structure of learning approaches to a temporal level. Here, we present a Dynamic BN framework, which learns interaction networks between different time points in time-series data. Both intra and inter networks are learnt and compared to standard DBN learning algorithms. Our results based on synthetic and simulated gene expression data suggest that the proposed method outperforms existing approaches in identifying the underlying network structure. The proposed framework is robust to errors in the incorporated knowledge and can combine various experimental data types together with existing knowledge when learning networks.

AB - One of the main problems in systems biology is learning gene interaction networks from experimental data. This turns out to be a challenging task as the experimental data is sparse and noisy, and network learning algorithms are computationally intense. Bayesian Networks (BN) have become a popular choice for learning such networks as BNs avoid overfitting and are robust to noise. In this paper we build up on our established framework, Bayesian Network Prior, where we incorporate existing biological knowledge in learning gene interaction networks. However, biological phenomena are time-dependent and there is need to extend the static structure of learning approaches to a temporal level. Here, we present a Dynamic BN framework, which learns interaction networks between different time points in time-series data. Both intra and inter networks are learnt and compared to standard DBN learning algorithms. Our results based on synthetic and simulated gene expression data suggest that the proposed method outperforms existing approaches in identifying the underlying network structure. The proposed framework is robust to errors in the incorporated knowledge and can combine various experimental data types together with existing knowledge when learning networks.

KW - Dynamic Bayesian Networks

KW - External biological knowledge

KW - Gene interaction networks

KW - Microarray

KW - Time-series data

UR - http://www.scopus.com/inward/record.url?scp=84892630819&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84892630819&partnerID=8YFLogxK

U2 - 10.1109/HIBIT.2013.6661680

DO - 10.1109/HIBIT.2013.6661680

M3 - Conference contribution

AN - SCOPUS:84892630819

SN - 9781479907014

T3 - 2013 8th International Symposium on Health Informatics and Bioinformatics, HIBIT 2013

BT - 2013 8th International Symposium on Health Informatics and Bioinformatics, HIBIT 2013

PB - IEEE Computer Society

ER -