Functional identification of cis-regulatory long noncoding RNAs at controlled false discovery rates

Nucleic Acids Research, Journal Year: 2024, Volume and Issue: 52(6), P. 2821 - 2835

Published: Feb. 13, 2024

Abstract A key attribute of some long noncoding RNAs (lncRNAs) is their ability to regulate expression neighbouring genes in cis. However, such ‘cis-lncRNAs’ are presently defined using ad hoc criteria that, we show, prone false-positive predictions. The resulting lack cis-lncRNA catalogues hinders our understanding extent, characteristics and mechanisms. Here, introduce TransCistor, a framework for defining identifying cis-lncRNAs based on enrichment targets amongst proximal genes. TransCistor’s simple conservative statistical models compatible with functionally target gene maps generated by existing future technologies. Using transcriptome-wide perturbation experiments 268 human 134 mouse lncRNAs, provide the first large-scale survey cis-lncRNAs. Known correctly identified, including XIST, LINC00240 UMLILO, predictions consistent across analysis methods, types independent experiments. We detect cis-activity minority primarily involving activators over repressors. Cis-lncRNAs detected both RNA interference antisense oligonucleotide perturbations. Mechanistically, transcripts observed physically associate weakly enriched enhancer elements. In summary, TransCistor establishes quantitative foundation cis-lncRNAs, opening path elucidating molecular mechanisms biological significance.

Language: Английский

Exploring the roles of RNAs in chromatin architecture using deep learning

Nature Communications, Journal Year: 2024, Volume and Issue: 15(1)

Published: July 29, 2024

Abstract Recent studies have highlighted the impact of both transcription and transcripts on 3D genome organization, particularly its dynamics. Here, we propose a deep learning framework, called AkitaR, that leverages sequences genome-wide RNA-DNA interactions to investigate roles chromatin-associated RNAs (caRNAs) folding in HFFc6 cells. In order disentangle cis - trans -regulatory caRNAs, compared models with nascent transcripts, -located open chromatin data, or DNA sequence alone. Both caRNAs improve models’ predictions, especially at cell-type-specific genomic regions. Analyses feature importance scores reveal contribution TAD boundaries, loops nuclear sub-structures such as speckles nucleoli predictions. Furthermore, identify non-coding (ncRNAs) known regulate structures, MALAT1 NEAT1 , well several new RNAs, RNY5 RPPH1 POLG-DT THBS1-IT1 might modulate architecture through -interactions HFFc6. Our modeling also suggests from Alus other repetitive elements may facilitate R-loop formation. findings provide insights generate testable hypotheses about shaping organization.

Language: Английский

Functional identification of cis-regulatory long noncoding RNAs at controlled false-discovery rates

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2022, Volume and Issue: unknown

Published: Sept. 19, 2022

ABSTRACT A key attribute of some long noncoding RNAs (lncRNAs) is their ability to regulate expression neighbouring genes in cis. However, such ‘cis-lncRNAs’ are presently defined using ad hoc criteria that, we show, prone false-positive predictions. The resulting lack cis-lncRNA catalogues hinders our understanding extent, characteristics and mechanisms. Here, introduce TransCistor, a framework for defining identifying cis-lncRNAs based on enrichment targets amongst proximal genes. TransCistor’s simple conservative statistical models compatible with functionally-defined target gene maps generated by existing future technologies. Using transcriptome-wide perturbation experiments 268 human 134 mouse lncRNAs, provide the first large-scale survey cis-lncRNAs. Known correctly identified, including XIST, LINC00240 UMLILO, predictions consistent across analysis methods, types independent experiments. Our results indicate that cis-activity detected minority primarily involving activators over repressors. Cis-lncRNAs both RNA interference antisense oligonucleotide perturbations. Mechanistically, transcripts observed physically associate target-genes, weakly enriched enhancer-elements. In summary, TransCistor establishes quantitative foundation cis-lncRNAs, opening path elucidating molecular mechanisms biological significance.

Language: Английский