Multi-level gene/MiRNA feature selection using deep belief nets and active learning

Rania Ibrahim, Noha Yousri, Mohamed A. Ismail, Nagwa M. El-Makky

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

27 Citations (Scopus)

Abstract

Selecting the most discriminative genes/miRNAs has been raised as an important task in bioinformatics to enhance disease classifiers and to mitigate the dimensionality curse problem. Original feature selection methods choose genes/miRNAs based on their individual features regardless of how they perform together. Considering group features instead of individual ones provides a better view for selecting the most informative genes/miRNAs. Recently, deep learning has proven its ability in representing the data in multiple levels of abstraction, allowing for better discrimination between different classes. However, the idea of using deep learning for feature selection is not widely used in the bioinformatics field yet. In this paper, a novel multi-level feature selection approach named MLFS is proposed for selecting genes/miRNAs based on expression profiles. The approach is based on both deep and active learning. Moreover, an extension to use the technique for miRNAs is presented by considering the biological relation between miRNAs and genes. Experimental results show that the approach was able to outperform classical feature selection methods in hepatocellular carcinoma (HCC) by 9%, lung cancer by 6% and breast cancer by around 10% in F1-measure. Results also show the enhancement in F1-measure of our approach over recently related work in [1] and [2].

Original languageEnglish
Title of host publication2014 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2014
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages3957-3960
Number of pages4
ISBN (Electronic)9781424479290
DOIs
Publication statusPublished - 1 Jan 2014
Externally publishedYes
Event2014 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2014 - Chicago, United States
Duration: 26 Aug 201430 Aug 2014

Other

Other2014 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2014
CountryUnited States
CityChicago
Period26/8/1430/8/14

Fingerprint

Problem-Based Learning
MicroRNAs
Feature extraction
Genes
Bioinformatics
Computational Biology
Learning
Breast Neoplasms
Aptitude
Classifiers
Hepatocellular Carcinoma
Lung Neoplasms
Deep learning

ASJC Scopus subject areas

  • Health Informatics
  • Computer Science Applications
  • Biomedical Engineering

Cite this

Ibrahim, R., Yousri, N., Ismail, M. A., & El-Makky, N. M. (2014). Multi-level gene/MiRNA feature selection using deep belief nets and active learning. In 2014 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2014 (pp. 3957-3960). [6944490] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/EMBC.2014.6944490

Multi-level gene/MiRNA feature selection using deep belief nets and active learning. / Ibrahim, Rania; Yousri, Noha; Ismail, Mohamed A.; El-Makky, Nagwa M.

2014 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2014. Institute of Electrical and Electronics Engineers Inc., 2014. p. 3957-3960 6944490.

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

Ibrahim, R, Yousri, N, Ismail, MA & El-Makky, NM 2014, Multi-level gene/MiRNA feature selection using deep belief nets and active learning. in 2014 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2014., 6944490, Institute of Electrical and Electronics Engineers Inc., pp. 3957-3960, 2014 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2014, Chicago, United States, 26/8/14. https://doi.org/10.1109/EMBC.2014.6944490
Ibrahim R, Yousri N, Ismail MA, El-Makky NM. Multi-level gene/MiRNA feature selection using deep belief nets and active learning. In 2014 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2014. Institute of Electrical and Electronics Engineers Inc. 2014. p. 3957-3960. 6944490 https://doi.org/10.1109/EMBC.2014.6944490
Ibrahim, Rania ; Yousri, Noha ; Ismail, Mohamed A. ; El-Makky, Nagwa M. / Multi-level gene/MiRNA feature selection using deep belief nets and active learning. 2014 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2014. Institute of Electrical and Electronics Engineers Inc., 2014. pp. 3957-3960
@inproceedings{153f9b0b809547f7a3fea82892beb9d1,
title = "Multi-level gene/MiRNA feature selection using deep belief nets and active learning",
abstract = "Selecting the most discriminative genes/miRNAs has been raised as an important task in bioinformatics to enhance disease classifiers and to mitigate the dimensionality curse problem. Original feature selection methods choose genes/miRNAs based on their individual features regardless of how they perform together. Considering group features instead of individual ones provides a better view for selecting the most informative genes/miRNAs. Recently, deep learning has proven its ability in representing the data in multiple levels of abstraction, allowing for better discrimination between different classes. However, the idea of using deep learning for feature selection is not widely used in the bioinformatics field yet. In this paper, a novel multi-level feature selection approach named MLFS is proposed for selecting genes/miRNAs based on expression profiles. The approach is based on both deep and active learning. Moreover, an extension to use the technique for miRNAs is presented by considering the biological relation between miRNAs and genes. Experimental results show that the approach was able to outperform classical feature selection methods in hepatocellular carcinoma (HCC) by 9{\%}, lung cancer by 6{\%} and breast cancer by around 10{\%} in F1-measure. Results also show the enhancement in F1-measure of our approach over recently related work in [1] and [2].",
author = "Rania Ibrahim and Noha Yousri and Ismail, {Mohamed A.} and El-Makky, {Nagwa M.}",
year = "2014",
month = "1",
day = "1",
doi = "10.1109/EMBC.2014.6944490",
language = "English",
pages = "3957--3960",
booktitle = "2014 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2014",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - GEN

T1 - Multi-level gene/MiRNA feature selection using deep belief nets and active learning

AU - Ibrahim, Rania

AU - Yousri, Noha

AU - Ismail, Mohamed A.

AU - El-Makky, Nagwa M.

PY - 2014/1/1

Y1 - 2014/1/1

N2 - Selecting the most discriminative genes/miRNAs has been raised as an important task in bioinformatics to enhance disease classifiers and to mitigate the dimensionality curse problem. Original feature selection methods choose genes/miRNAs based on their individual features regardless of how they perform together. Considering group features instead of individual ones provides a better view for selecting the most informative genes/miRNAs. Recently, deep learning has proven its ability in representing the data in multiple levels of abstraction, allowing for better discrimination between different classes. However, the idea of using deep learning for feature selection is not widely used in the bioinformatics field yet. In this paper, a novel multi-level feature selection approach named MLFS is proposed for selecting genes/miRNAs based on expression profiles. The approach is based on both deep and active learning. Moreover, an extension to use the technique for miRNAs is presented by considering the biological relation between miRNAs and genes. Experimental results show that the approach was able to outperform classical feature selection methods in hepatocellular carcinoma (HCC) by 9%, lung cancer by 6% and breast cancer by around 10% in F1-measure. Results also show the enhancement in F1-measure of our approach over recently related work in [1] and [2].

AB - Selecting the most discriminative genes/miRNAs has been raised as an important task in bioinformatics to enhance disease classifiers and to mitigate the dimensionality curse problem. Original feature selection methods choose genes/miRNAs based on their individual features regardless of how they perform together. Considering group features instead of individual ones provides a better view for selecting the most informative genes/miRNAs. Recently, deep learning has proven its ability in representing the data in multiple levels of abstraction, allowing for better discrimination between different classes. However, the idea of using deep learning for feature selection is not widely used in the bioinformatics field yet. In this paper, a novel multi-level feature selection approach named MLFS is proposed for selecting genes/miRNAs based on expression profiles. The approach is based on both deep and active learning. Moreover, an extension to use the technique for miRNAs is presented by considering the biological relation between miRNAs and genes. Experimental results show that the approach was able to outperform classical feature selection methods in hepatocellular carcinoma (HCC) by 9%, lung cancer by 6% and breast cancer by around 10% in F1-measure. Results also show the enhancement in F1-measure of our approach over recently related work in [1] and [2].

UR - http://www.scopus.com/inward/record.url?scp=84929492596&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84929492596&partnerID=8YFLogxK

U2 - 10.1109/EMBC.2014.6944490

DO - 10.1109/EMBC.2014.6944490

M3 - Conference contribution

C2 - 25570858

AN - SCOPUS:84929492596

SP - 3957

EP - 3960

BT - 2014 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2014

PB - Institute of Electrical and Electronics Engineers Inc.

ER -