Project description:Oxford Nanopore enables direct RNA sequencing allowing for base calling of RNA modifications. We tested mouse hippocampi RNA samples, using a Nanopore direct RNA-seq protocol that in addition to long poly A selected RNAs allows sequencing also of non-poly A RNAs as well as short RNAs < 200nt (including SINE B2 RNAs and other non poly A non coding RNAs). We provide here as a resource a direct RNA-sequencing dataset generated from mouse brain tissues that includes both mRNAs and non poly A or short non-coding RNAs such as SINEs. Elevated SINE B2 RNA Adenosine to Inosine editing is consistently observed across hippocampal tissues of a mouse model of amyloid beta accumulation compared to hippocampi of wild type animals. Nanopore direct RNA sequencing supports increased RNA modification signals at the sameSINE B2 RNA regions identified by short-read Illumina sequencing in these hippocampi.
Project description:To detect the modifed bases in SINEUP RNA, we compared chemically modified in vitro transcribed (IVT) SINEUP-GFP RNA and in-cell transcribed (ICT) SINEUP RNA from SINEUP-GFP and sense EGFP co-transfected HEK293T/17 cells. Comparative study of Nanopore direct RNA sequencing data from non-modified and modified IVT samples against the data from ICT SINEUP RNA sample revealed modified k-mers positions in SINEUP RNA in the cell.
Project description:State-of-the-art algorithms for m6A detection and quantification via nanopore direct RNA sequencing have been continuously developed, little is known about their capacities and limitations, which makes a comprehensive assessment in urgent need. Therefore, we performed comprehensive benchmarking of 10 computational tools relying on current-based and base-calling “errors” strategies for m6A detection by nanopore sequencing.
Project description:Long-read nanopore sequencing has emerged as a potent tool for studying RNA modifications. However, the detection of N4-acetylcytidine (ac4C) based on nanopore sequencing remains largely unexplored. Here, we introduce ac4Cnet, a deep learning frame utilizing Oxford Nanopore direct RNA sequencing to accurately identify ac4C sites. Our methodology involves training ac4Cnet capable of distinguishing ac4C from unmodified cytidine and 5-methylcytosine (m5C), as well as estimating the modification rate at each ac4C site. We demonstrate the robustness of our approach through validations on independent in vitro datasets and a human cell line, highlighting its versatility and potential for advancing the study of ac4C modifications.