Dataset Information

TRForest: a novel random forest-based algorithm for tRNA-derived fragment target prediction

ABSTRACT: tRNA fragments (tRFs) are a novel class of small RNAs comparable to the size and function of miRNAs. We and others have shown that tRFs are generally Dicer independent, can be found in abundance in the miRNA effector protein Ago, and can repress expression of specific genes that have complementarity to their 5’ seed-sequences. Given that this greatly expands the repertoire of small RNAs capable of post-transcriptional gene expression, it is important to predict tRF targets with confidence. Some attempts have been made to predict tRF targets, but are limited in the scope of tRF classes used in prediction or limited in feature selection. We hypothesized that established miRNA target prediction features applied to tRFs through a random forest machine learning algorithm will immensely improve tRF target prediction. Using this approach, we show significant improvements in tRF target prediction for all classes of tRFs and validate our predictions in two independent cell lines. Finally, using Gene Ontology analysis, we provide evidence that tRF-3009a targets may be involved in neural development. These improvements to tRF target prediction further our understanding of tRF function broadly across species and tRF types, and provide avenues for testing novel roles for tRFs in biology.

ORGANISM(S): Homo sapiens

PROVIDER: GSE189510 | GEO | 2022/05/16

REPOSITORIES: GEO

ACCESS DATA

Dataset's files

Source:

			Action	DRS
		Other

Items per page:

1 - 1 of 1

Similar Datasets

Project description:Background: As couples struggle with infertility and livestock producers wish to rapidly improve genetic merit in their herd, assisted reproductive technologies (ART) have become increasingly popular in human medicine as well as the livestock industry. Utilizing ART can cause an increased risk of congenital overgrowth syndromes, such as Large Offspring Syndrome (LOS) in ruminants. A dysregulation of transcripts has been observed in bovine fetuses with LOS, which is suggested to be a cause of the phenotype. Our recent study identified variations in tRNA expression in LOS individuals, leading us to hypothesize that variations in tRNA expression can influence the availability of their processed regulatory products, tRNA-derived fragments (tRFs). Due to their resemblance in size to microRNAs, studies suggest that tRFs target mRNA transcripts and regulate gene expression. Thus, we have sequenced small RNA isolated from skeletal muscle and liver of day 105 bovine fetuses to elucidate the mechanisms contributing to LOS. Moreover, we have utilized our previously generated tRNA sequencing data to analyze the contribution of tRNA availability to tRF abundance. Results: 22,289 and 7,737 unique tRFs were predicted in the liver and muscle tissue respectively. The greatest number of reads originated from 5′ tRFs in muscle and 5′ halves in liver. In addition, mitochondrial (MT) and nuclear derived tRF expression was tissue-specific with most MT-tRFs and nuclear tRFs derived from LysUUU and iMetCAU in muscle, and AsnGUU and GlyGCC in liver. Despite variation in tRF abundance within treatment groups, we identified differentially expressed (DE) tRFs across Control-AI, ART-Normal, and ART-LOS groups with the most DE tRFs between ART-Normal and ART-LOS groups. Many DE tRFs target transcripts enriched in pathways related to growth and development in the muscle and tumor development in the liver. Finally, we found positive correlation coefficients between tRNA availability and tRF expression in muscle (R = 0.47) and liver (0.6). Conclusion: Our results highlight the dysregulation of tRF expression and its regulatory roles in LOS. These tRFs were found to target both imprinted and non-imprinted genes in muscle as well as genes linked to tumor development in the liver. Furthermore, we found that tRNA transcription is a highly modulated event that plays a part in the biogenesis of tRFs. This study is the first to investigate the relationship between tRNA and tRF expression in combination with ART-induced LOS.

Dataset Information

TRForest: a novel random forest-based algorithm for tRNA-derived fragment target prediction

Dataset's files

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets