<i>microPred</i>: effective classification of pre-miRNAs for human miRNA gene prediction

Batuwita, Rukshan; Palade, Vasile

doi:10.1093/bioinformatics/btp107

Cited by 238 publications

(249 citation statements)

References 35 publications

Supporting

Mentioning

247

Contrasting

Order By: Relevance

“…For these methods the classification process of hairpins is transparent, which is rarely the case for machine learning methods. Triplet-SVM (Xue et al, 2005), microPred (Batuwita and Palade, 2009), MiRPara (Wu et al, 2011), Mirident (Liu et al, 2012) all rely on SVMs, but differ with respect to the hairpin features included. Triplet-SVM is a relatively simple model that relies on that the propensity of secondary structure triplets within the hairpin structure being different between miRNAs and pseudo hairpins.…”

Section: Hairpin Classification Methodsmentioning

confidence: 99%

Genome-wide discovery of miRNAs using ensembles of machine learning algorithms and logistic regression

Ulfenborg

Klinga-Levan

Olsson

2015

IJDMB

View full text Add to dashboard Cite

Abstract:In silico prediction of novel miRNAs from genomic sequences remains a challenging problem. This study presents a genome-wide miRNA discovery software package called GenoScan and evaluates two hairpin classification methods. These methods, one ensemble-based and one using logistic regression were benchmarked along with 15 published methods. In addition, the sequence-folding step is addressed by investigating the impact of secondary structure prediction methods and the choice of input sequence length on prediction performance. Both the accuracy of secondary structure predictions and the miRNA prediction are evaluated. In the benchmark of hairpin classification methods, the regression model achieved highest classification accuracy. Of the structure prediction methods evaluated, ContextFold achieved the highest agreement between predicted and experimentally determined structures. However, both the choice of secondary structure prediction method and input sequence length had limited impact on hairpin classification performance.

show abstract

Section: Hairpin Classification Methodsmentioning

confidence: 99%

Genome-wide discovery of miRNAs using ensembles of machine learning algorithms and logistic regression

Ulfenborg

Klinga-Levan

Olsson

2015

IJDMB

View full text Add to dashboard Cite

show abstract

“…In medical science, bioinformatics, and machine learning communities [23,24,33,34], the sensitivity (SE) and the specificity (SP) are two metrics used to evaluate the performance of classifiers. Sensitivity measures the proportion of actual positives which are correctly identified as such, while specificity can be defined as the proportion of negatives which are correctly identified.…”

Section: Evaluation Measuresmentioning

confidence: 99%

“…In addition, miRNA prediction problem should be also considered as imbalance class distribution. Batuwita et al [23] proposed an effective classifier system, namely microPred which used a complete pseudo hairpin dataset and ncRNAs. In their research, they focused on handling with class imbalance problem in datasets where samples from the majority class (9248 = 8494 pseudo hairpins + 754 other ncRNAs) significantly outnumber the minority class (691 pre-miRNAs).…”

Section: Introductionmentioning

confidence: 99%

“…One of remarkable solutions is SMOTE which is a famous and general over-sampling method addressing this problem. For example, experimental results of Batuwita et al [23] suggested that the best classifier has been developed by applying the SMOTE method. However, there are still some drawbacks of SMOTE, particularly in some cases it cannot improve or sometimes reduces classification performance.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A novel over-sampling method and its application to miRNA prediction

Dang¹,

Hirose²,

Saethang³

et al. 2013

JBiSE

View full text Add to dashboard Cite

MicroRNAs (miRNAs) are short (~22 nt) non-coding RNAs that play an indispensable role in gene regulation of many biological processes. Most of current computational, comparative, and non-comparative methods commonly classify human precursor micro-RNA (pre-miRNA) hairpins from both genome pseudo hairpins and other non-coding RNAs (ncRNAs). Although there were a few approaches achieving promising results in applying class imbalance learning methods, this issue has still not solved completely and successfully yet by the existing methods because of imbalanced class distribution in the datasets. For example, SMOTE is a famous and general over-sampling method addressing this problem, however in some cases it cannot improve or sometimes reduces classification performance. Therefore, we developed a novel over-sampling method named incre-mental-SMOTE to distinguish human pre-miRNA hairpins from both genome pseudo hairpins and other ncRNAs. Experimental results on pre-miRNA datasets from Batuwita et al. showed that our method achieved better Sensitivity and G-mean than the control (no oversampling), SMOTE, and several successsors of modified SMOTE including safe-level-SMOTE and border-line-SMOTE. In addition, we also applied the novel method to five imbalanced benchmark datasets from UCI Machine Learning Repository and achieved improvements in Sensitivity and G-mean. These results suggest that our method outperforms SMOTE and several successors of it in various biomedical classification problems including miRNA classification.

show abstract

“…microPred microPred presents an effective classifier with appropriate machine learning techniques. The approach in microPred includes the introduction of more representative datasets, extraction of new biologically relevant features, feature selection, handling of class imbalance problems in datasets, and extensive classifier performance evaluation via systematic crossvalidation methods (Batuwita et al, 2009). …”

Section: Mipredmentioning

confidence: 99%

Benchmark comparison of ab initio microRNA identification methods and software

Huang²,

Wang³

et al. 2012

Genet. Mol. Res.

View full text Add to dashboard Cite

ABSTRACT. MicroRNAs (miRNAs) are short, non-coding RNA molecules that play an important role in the world of genes, especially in regulating the gene expression of target messenger RNAs through cleavage or translational repression of messenger RNA. Ab initio methods have become popular in computational miRNA detection. Most software tools are designed to distinguish miRNA precursors from pseudo-hairpins, but a few can mine miRNA from genome or expressed sequence tag sequences. We prepared novel testing datasets to measure and compare the performance of various software tools. Furthermore, we summarized the miRNA mining methods that study next-generation sequencing data for bioinformatics researchers who are analyzing these data. Because secondary structure is an important feature in the identification of miRNA, we analyzed the influence of various secondary structure prediction software tools on miRNA identification. MiPred was the most effective for classifying real-/ pseudo-pre-miRNA sequences, and miRAbela performed relatively ©FUNPEC-RP www.funpecrp.com.br Genetics and Molecular Research 11 (4): 4525-4538 (2012) L.L. Hu et al. 4526 better for mining miRNA precursors from genome or expressed sequence tag sequences. RNA-fold performed better than m-fold for extracting secondary structure features of miRNA precursors.

show abstract

microPred: effective classification of pre-miRNAs for human miRNA gene prediction

Cited by 238 publications

References 35 publications

Genome-wide discovery of miRNAs using ensembles of machine learning algorithms and logistic regression

Genome-wide discovery of miRNAs using ensembles of machine learning algorithms and logistic regression

A novel over-sampling method and its application to miRNA prediction

Benchmark comparison of ab initio microRNA identification methods and software

Contact Info

Product

Resources

About