Dataset Information

CoRAL: predicting non-coding RNAs from small RNA-sequencing data.

ABSTRACT: The surprising observation that virtually the entire human genome is transcribed means we know little about the function of many emerging classes of RNAs, except their astounding diversities. Traditional RNA function prediction methods rely on sequence or alignment information, which are limited in their abilities to classify the various collections of non-coding RNAs (ncRNAs). To address this, we developed Classification of RNAs by Analysis of Length (CoRAL), a machine learning-based approach for classification of RNA molecules. CoRAL uses biologically interpretable features including fragment length and cleavage specificity to distinguish between different ncRNA populations. We evaluated CoRAL using genome-wide small RNA sequencing data sets from four human tissue types and were able to classify six different types of RNAs with ?80% cross-validation accuracy. Analysis by CoRAL revealed that microRNAs, small nucleolar and transposon-derived RNAs are highly discernible and consistent across all human tissue types assessed, whereas long intergenic ncRNAs, small cytoplasmic RNAs and small nuclear RNAs show less consistent patterns. The ability to reliably annotate loci across tissue types demonstrates the potential of CoRAL to characterize ncRNAs using small RNA sequencing data in less well-characterized organisms.

SUBMITTER: Leung YY

PROVIDER: S-EPMC3737537 | biostudies-literature | 2013 Aug

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

CoRAL: predicting non-coding RNAs from small RNA-sequencing data.

Leung Yuk Yee YY Ryvkin Paul P Ungar Lyle H LH Gregory Brian D BD Wang Li-San LS

Nucleic acids research 20130521 14

The surprising observation that virtually the entire human genome is transcribed means we know little about the function of many emerging classes of RNAs, except their astounding diversities. Traditional RNA function prediction methods rely on sequence or alignment information, which are limited in their abilities to classify the various collections of non-coding RNAs (ncRNAs). To address this, we developed Classification of RNAs by Analysis of Length (CoRAL), a machine learning-based approach f ...[more]

PMID: 23700308

Similar Datasets

Project description:Background: Improved healthcare and recent breakthroughs in technology have substantially reduced cancer mortality rates worldwide. Recent advancements in next-generation sequencing (NGS) have allowed genomic analysis of the human transcriptome. Now, using NGS we can further look into small non-coding regions of RNAs (sncRNAs) such as microRNAs (miRNAs), Piwi-interacting-RNAs (piRNAs), long non-coding RNAs (lncRNAs), and small nuclear/nucleolar RNAs (sn/snoRNAs) among others. Recent studies looking at sncRNAs indicate their role in important biological processes such as cancer progression and predict their role as biomarkers for disease diagnosis, prognosis, and therapy. Results: In the present study, we data mined publically available small RNA sequencing data from colorectal tissue samples of eight matched patients (benign, tumor, and metastasis) and remapped the data for various small RNA annotations. We identified aberrant expression of 13 miRNAs in tumor and metastasis specimens [tumor vs benign group (19 miRNAs) and metastasis vs benign group (38 miRNAs)] of which five were upregulated, and eight were downregulated, during disease progression. Pathway analysis of aberrantly expressed miRNAs showed that the majority of miRNAs involved in colon cancer were also involved in other cancers. Analysis of piRNAs revealed six to be over-expressed in the tumor vs benign cohort and 24 in the metastasis vs benign group. Only two piRNAs were shared between the two cohorts. Examining other types of small RNAs [sn/snoRNAs, mt_rRNA, miscRNA, nonsense mediated decay (NMD), and rRNAs] identified 15 sncRNAs in the tumor vs benign group and 104 in the metastasis vs benign group, with only four others being commonly expressed. Conclusion: In summary, our comprehensive analysis on publicly available small RNA-seq data identified multiple differentially expressed sncRNAs during colorectal cancer progression at different stages compared to normal colon tissue. We speculate that deciphering and validating the roles of sncRNAs may prove useful in colorectal cancer prognosis, diagnosis, and therapy.

Project description:PIWI-interacting RNAs (piRNAs) are a novel class of small ncRNAs initially isolated from germ line cells; although recent studies report that they are expressed also in somatic cells. To elucidate the role of piRNAs in somatic cells, in particular from breast cancer, we performed the first extensive next generation sequencing expression analysis of small RNA transcriptomes of hormone responsive breast cancer cell lines in different culture conditions. In addition, to understand the behavior of piRNAs with respect to miRNAs in breast tumor tissues, small RNA sequence data set available in Gene Expression Omnibus (GSE39162) database was used. Results led to the identification of ~100 and ~150 human piRNAs in the breast cancerous cell lines and tumors respectively, several of which differentially expressed in cell lines under different experimental conditions tested or in response to ERβ and in tumor tissues. Western blotting and Q-PCR analysis also revealed the presence in breast cell lines of PIWIL (PIWI Like) subfamily members proteins encoded in the human genome (PIWIL2/HILI and PIWIL4/HIWI2) and of other components of the piRNA biogenesis pathways, suggesting that this might indeed be functional in somatic cells. These results show that piRNAs are expressed in human somatic cells, in particular in cancer, where their expression is influenced by neoplastic transformation, growth conditions and estrogen receptor beta. More important, we demonstrate for the first time a distinct pattern of piRNAs expression in cancerous vs normal breast tissues, which suggests a potential role of these epigenetic modulators in mammary carcinogenesis and maintenance of the cancer cell phenotype. In addition, to understand the behavior of piRNAs with respect to miRNAs in breast tumor tissues, small RNA sequence data set available in Gene Expression Omnibus (GEO; GSM957192 TAX577740T ,GSM957194 TAX577740N, GSM957195 TAX577453T, GSM957197 TAX577453N, GSM957198 TAX577745T, GSM957200 TAX577745N, GSM957201 TAX577579T, GSM957203 TAX577579N) was used.

Dataset Information

CoRAL: predicting non-coding RNAs from small RNA-sequencing data.

Publications

CoRAL: predicting non-coding RNAs from small RNA-sequencing data.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets