Bin Xue
ABSTRACT
MiRNAs are small non-coding RNAs, but perform important function in regulating the translation of mRNAs into proteins. Experimental identification of miRNAs is still challenging due to the fact that the biogenesis of miRNAs is a complex process and is regulated by many factors that have not been completely characterized. Therefore, developing knowledge-based computational predictors provides an alternative strategy to discover novel miRNAs. While many computational predictors have been developed, a perpetual theme in this field is to improve the prediction performance continuously. Inspired by the success of deep leaning strategies in many other fields, we developed novel miRNA meta-predictors by integrating several deep learning techniques, including: preprocessing of input features based on principal component analysis and non-linear transformation, and meta-strategy. The newly designed meta-predictors improved the prediction accuracy of miRNAs consistently by ~10% from the middle of 80% to the middle of 90%.
- Kozomara, A. and Griffiths-Jones, S. 2014. miRBase: annotating high confidence microRNAs using deep sequencing data. Nucl. Acids. Res. 42, D1 (Jan. 2014), D68--D73.Google Scholar
Cross Ref
- Lewis, B. P., Burge, C. B., and Bartel, D. P. 2005. Conserved seed pairing, often flanked by adenosines, indicates that thousands of human genes are microRNAs targets. Cell. 120, 1 (Jan. 2005), 15--20.Google ScholarCross Ref
- Friedman, R. C., Farh, K. K., Burge, C. B., and Bartel, D. P. 2009. Most mammalian mRNAs are conserved targets of microRNAs. Genome Res.19, 1 (Jan. 2009), 92--105.Google Scholar
- Xue, B., Dor, O., Faraggi, E., and Zhou Y. 2008. Real-value prediction of backbone torsion angles. Proteins.72, 1 (Jul. 2008), 427--433.Google ScholarCross Ref
- Xue, B., Faraggi, E., and Zhou Y. 2009. Predicting residue-residue contact map by a two-layer, integrated neural-network method. Proteins.76, 1 (Jul. 2009), 176--183.Google ScholarCross Ref
- Xue, B., Oldfield, C. J., Dunker, A. K., and Uversky, V. N. 2009. CDF it all: consensus prediction of intrinsically disordered proteins based on various cumulative distribution functions. FEBS Lett.583, 9 (May. 2009), 1469--1474.Google ScholarCross Ref
- Xue, B., Dunbrack, R. L., Williams, R. W., Dunker, A. K., and Uversky, V. N. 2010. PONDR-FIT: a meta-predictor of intrinsically disordered amino acids. Biochim. Biophys. Acta..1804, 4 (Apr. 2010), 996--1010.Google Scholar
- Jiang, P., Wu. H., Wang, W., Ma. W., Sun, X., and Lu. Z. 2007. MiPred: classification of real and pseudo microRNAs precursors using random forest prediction model with combined features. Nucl. Acids. Res.35, (Jul. 2007), W339--W344.Google ScholarCross Ref
- Mathelier A. and Carbone A. 2010. MIReNA: finding microRNAs with high accuracy and no learning at genome scale and from deep sequencing data. Bioinformatics.26, 18 (Sep. 2010), 2226--2234. Google ScholarDigital Library
- Xue, C., Li, F., He, T., Liu, G. P., Li, Y., and Zhang X. 2005. Classification of real and pseudo microRNA precursors using local structure-sequence features and support vector machine. BMC Bioinformatics.6, (Dec. 2005), 310.Google Scholar
- Wu, Y., Wei, B., Liu, H., Li, T., and Rayner S. 2011. MiRPara: a SVM-based software tool for prediction of most probable microRNA coding regions in genome scale sequences. BMC Bioinformatics.12, (Apr. 2011), 107.Google Scholar
- Nam, J. W., Kim, J., Kim, S. K., and Zhang, B. T. 2006. ProMiR II: a web server for the probabilistic prediction of clustered, nonclustered, conserved and nonconserved microRNAs. Nucl. Acids. Res.34, (Jul. 2006), W455--W458.Google Scholar
- Nawrocki, E. P., Burge, S. W., Bateman, A., Daub, J., Eberhardt, R. Y., Eddy, S. R., Floden, E. W. Gardner, P. P., Jones, T. A., Tate, J., and Finn, R. D. 2015. Rfam 12.0: updates to the RNA families database. Nucl. Acids. Res.43, (Jan. 2015), D130--D137.Google Scholar
Index Terms
- Improving MiRNA prediction accuracy by deep learning strategies
Recommendations
Discovery of functional miRNA-mRNA regulatory modules with computational methods
The identification of miRNAs and their target mRNAs and the construction of their regulatory networks may give new insights into biological procedures. This study proposes a computational method to discover the functional miRNA-mRNA regulatory modules (...
A deep learning framework for identification of microRNA regulatory modules: student research abstract
SAC '17: Proceedings of the Symposium on Applied ComputingMicroRNAs (miRNAs) belong to the family of RNAs and are known to repress the expressions of their targets. They work cooperatively with genes in post-transcriptional gene regulation called miRNA regulatory modules (MRMs). Furthermore, understanding the ...
Inferring functional miRNA-mRNA regulatory modules in epithelial-mesenchymal transition with a probabilistic topic model
MicroRNAs (miRNAs) play important roles in gene regulatory networks. In this paper, we propose a probabilistic topic model to infer regulatory networks of miRNAs and their target mRNAs for specific biological conditions at the post-transcriptional level,...
Comments