DNA methylation patterns facilitate the identification of microRNA transcription start sites: a brain-specific study

Predicting the transcription start sites (TSSs) of microRNAs (miRNAs) is important for understanding how these small RNA molecules, known to regulate translation and stability of protein-coding genes, are regulated themselves. Previous approaches are primarily based on genetic features, trained on T...

Full description

Saved in:
Bibliographic Details
Published in:PloS one 2013-06, Vol.8 (6), p.e66722-e66722
Main Authors: Bhadra, Tapas, Bhattacharyya, Malay, Feuerbach, Lars, Lengauer, Thomas, Bandyopadhyay, Sanghamitra
Format: Article
Language:English
Subjects:
Algorithms
Bioinformatics
Biology
Brain
Coding
CpG islands
Deoxyribonucleic acid
DNA
DNA Methylation
Enzymes
Epigenesis, Genetic
Epigenetics
Genes
Genomes
Genomics
Humans
Identification
Mathematical models
MicroRNAs
MicroRNAs - genetics
miRNA
Model accuracy
Prediction models
Ribonucleic acid
RNA
Statistical analysis
Statistical models
Studies
Support Vector Machine
Transcription
Transcription, Genetic
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Predicting the transcription start sites (TSSs) of microRNAs (miRNAs) is important for understanding how these small RNA molecules, known to regulate translation and stability of protein-coding genes, are regulated themselves. Previous approaches are primarily based on genetic features, trained on TSSs of protein-coding genes, and have low prediction accuracy. Recently, a support vector machine based technique has been proposed for miRNA TSS prediction that uses known miRNA TSS for training the classifier along with a set of existing and novel CpG island based features. Current progress in epigenetics research has provided genomewide and tissue-specific reports about various phenotypic traits. We hypothesize that incorporating epigenetic characteristics into statistical models may lead to better prediction of primary transcripts of human miRNAs. In this paper, we have tested our hypothesis on brain-specific miRNAs by using epigenetic as well as genetic features to predict the primary transcripts. For this, we have used a sophisticated feature selection technique and a robust classification model. Our prediction model achieves an accuracy of more than 80% and establishes the potential of epigenetic analysis for in silico prediction of TSSs.
ISSN:1932-6203
1932-6203
DOI:10.1371/journal.pone.0066722