Development of sequencing methods to illuminate the epitranscriptome

• Main Author: Heng, Joel Jia Wei
• Other Authors: Tan Meng How
• Format: Thesis-Doctor of Philosophy
• Language: English
• Published: Nanyang Technological University 2025
• Subjects: Medicine, Health and Life Sciences; RNA; Modifications; Sequencing; Editing; Methylation
• Online Access: https://hdl.handle.net/10356/182139
• Institution: Nanyang Technological University

RNA modifications are chemical marks which expand genetic vocabulary beyond the canonical A, G, C, and T/U. The first RNA modification was discovered in 1957. While over 160 varieties of RNA modifications have been recorded since, the discovery that certain RNA modifications are reversible and can be erased sparked new enthusiasm in the research area. The improvements in sequencing technologies over the years also catalysed this renewed interest. Advancements in high-throughput sequencing have reduced costs and improved the accessibility of studies on the Epitranscriptome – RNA modifications and their associated RNA-binding proteins. The advent of nanopore sequencing further stoked the field’s interest with the introduction of direct RNA sequencing (DRS) which can sequence RNAs in their native state. DRS can be performed without reverse transcription, which typically erases the chemical modification of interest, and amplification, which potentially introduces biases. While showing much promise with direct RNA modification detection, DRS inherently contains a high noise background and requires more development. In this work, various sequencing technologies are employed to study the Epitranscriptome. RNA editing of Adenosine-to-Inosine (A to I) was catalogued and characterised in the model organism Xenopus with Illumina RNA-seq, revealing an interesting lack of conservation both between Xenopus species and between Xenopus and mammals. The editing sites discovered in Xenopus was subsequently used as part of the dataset to develop a deep learning-based method to perform transcriptome-wide detection of inosines in DRS sequencing data. Finally, ongoing efforts to develop a model that can identify different RNA modifications at specific bases on specific RNA molecules are discussed.