MicroRNA Prediction Based on Sample Classification Imbalance

Abstract:

MicroRNAs (miRNAs) play important regulatory roles in animals and plants by targeting mRNA for cleavage or translational repression. The main methods of predicting miRNA are biological experimental approaches and computational approaches. MiRNAs that have very low expression levels or that are expressed at specific stage are difficult to find by biological experiments. Computational approaches, especially machine learning approaches, can effectively overcome these difficulties. SVM (Support vector machine), which is one of the effective machine learning approaches, has a good performance on miRNA prediction. At present, the number of miRNA precursors that are experimentally validated is limited; however, the number of the sequence segment, which is similar to real miRNA precursors, is up to millions and millions. This caused classification imbalance when the samples are learned in the SVM. In this paper, the authors applied ensemble learning to solve this problem and achieve satisfactory performance.

• Periodical: Applied Mechanics and Materials (Volume 577)
• Pages: 1252-1257
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.577.1252
• Online since: July 2014
• Authors: Ji Ping Yang*
• Keywords: Classification Imbalance, Ensemble Learning, Machine Learning, MicroRNA, SVM

References

[1] E. Berezikov, E. Cuppen, R.H.A. Plasterk. Approaches to microRNA discovery[J]. Nature Genetics, 2006, 38(6s): S2-S7.

DOI: 10.1038/ng1794

[2] V.N. Kim,J. -W. Nam. Genomics of microRNA[J]. Trends in Genetics, 2006, 22(3): 165-173.

[3] C.H. Xue, F. Li, T. He, et al. Classification of real and pseudo microRNA precursors using local structure-sequence features and support vector machine[J]. BMC Bioinformatics, 2005, 6(1): 310-316.

DOI: 10.1186/1471-2105-6-310

[4] P. Jiang, H. Wu, W. Wang, et al. MiPred: classification of real and pseudo microRNA precursors using random forest prediction model with combined features[J]. Nucleic Acids Research, 2007, 35(Web Server): W339-W344.

DOI: 10.1093/nar/gkm368

[5] R. Batuwita,V. Palade. microPred: effective classification of pre-miRNAs for human miRNA gene prediction[J]. Bioinformatics, 2009, 25(8): 989-995.

DOI: 10.1093/bioinformatics/btp107

[6] S. Griffiths-Jones, H.K. Saini, S. van Dongen, et al. miRBase: tools for microRNA genomics[J]. Nucleic Acids Research, 2007, 36(Database): D154-D158.

DOI: 10.1093/nar/gkm952

[7] K.L.S. Ng,S.K. Mishra. De novo SVM classification of precursor microRNAs from genomic pseudo hairpins using global and intrinsic folding measures[J]. Bioinformatics, 2007, 23(11): 1321-1330.

DOI: 10.1093/bioinformatics/btm026

[8] Y. -W. Chen,C. -J. Lin. Combining SVMs with Various Feature Selection Strategies[J]. http: /www. csie. ntu. edu. tw/~cjlin/papers/features. pdf, (2009).

[9] C. -C. Chang,C. -J. Lin. LIBSVM: A Library for Support Vector Machines[D]. (2012).

[10] H. -Y. Wang, H. -K. Fan, Z. -A. Yao. et al. Imbalance dataset classification study[J]. 25, 2008, 5: 1301-1303.

[11] J. -G. Sun. Clustering Al gorithms Research[J]. Journal of Software, 2008, 19(1): 48-61.