Speakers – April 2017

Computational Identification of RNA Modifications

Kar Tong TAN
Research Assistant
Yang Lab
Cancer Science Institute
National University of Singapore

The epitranscriptome is an emerging layer of RNA regulation and is characterized by over 100 different RNA modifications that are known to exist on different RNA species. As compared to the relatively more well-known RNA modification events like A to I and C to U RNA editing, there is emerging evidence to indicate that mRNAs also carry other RNA modifications like m6A, pseudouridine, m5C and m1A which play important roles in the regulation of embryonic stem cell pluripotency and stress response. Nonetheless, due to the novelty of the RNA modification events, there is currently only very limited number of methods to identify the large variety of RNA modification events in the transcriptome. To add on, these methodologies often also require extremely specialized experimental protocols, which can be difficult to establish. Here, we propose a computational method for the identification of RNA modification events, which allows RNA modifications to be identified from RNA sequencing datasets with relative ease. Through the use of a specialized statistical model and stringent filtering criteria, sites of RNA modifications were identified. RNA modifications were also found to produce a distinct and unique signature thereby providing a unique feature for the classification of putative RNA modification events. Results using simulated datasets also confirmed the relatively high sensitivity and specificity of our method. Overall, we believe that the computational tool we established will ease the study of RNA modifications and enable its study in a wide variety of contexts.

Genome-wide identification of RNA sensors in prokaryotes and eukaryotes

Graduate Student
Wan Lab
Genome Institute of Singapore

Ability of RNA to fold into complex secondary and tertiary structures makes it suitable as cellular sensors on detection and response to metabolite changes in the environment (Breaker, 2010). Currently, RNA sensors are systematically identified through computational determination of sequence and structure homologies to known riboswitches (Barrick and Breaker, 2007). However, current established methods have limited capacity to reflect the diversity and distribution of RNA sensors in prokaryotes and eukaryotes as RNA can fold in different ways to bind to the same ligand, and sequence content can vary greatly between organisms (Wan et al., 2011). Here, we established a novel genome-wide method called “Parallel Analysis of RNA Conformations Exposed to Ligand binding (PARCEL)” specifically through the identification of RNA sensors directly and experimentally in both prokaryotic and eukaryotic genomes. Applying PARCEL to a collection of prokaryotic and eukaryotic organisms revealed new RNA sensors, which expanded the list of known RNA sensors to key metabolites. Conformable with their functional roles, the newly identified RNA sensors exhibit significant sequence and structure conservation, as well as ligand and species specificity. This collection of RNA sensors revolutionizes our understanding of their prevalence and distribution, and suggests that RNA based sensing and gene regulation is much more widespread than previously appreciated.


Barrick, J.E., and Breaker, R.R. (2007). The distributions, mechanisms, and structures of metabolite-binding riboswitches. Genome Biol. 8, R239.

Breaker, R.R. (2010). Riboswitches and the RNA World. Cold Spring Harb Perspect. Biol.

Wan, Y., Kertesz, M., Spitale, R.C., Segal, E., and Chang, H.Y. (2011). Understanding the transcriptome through RNA structure. Nat. Rev. Genet. 12, 641-655.

The Hidden Rule in ssRNA-Argonaute Protein Loading: Sequence Specificity

Eling GOH
Graduate Student
Okamura Lab
Nanyang Technological University

Argonuate (AGO) proteins play an important role in gene regulation with small RNAs (microRNAs and small interfering RNAs) serving as guides to targets. While guide strand in fly AGO2 has to be fully complementary to its target, base pairing at the seed region (2nd – 8th nucleotide of guide strand) is sufficient for fly AGO1. The duplex structure and 5’ terminal nucleotide identity are the only known determinants of AGO loading efficacy, and AGOs are believed to bind small RNAs independently of their sequences. However, we recently discovered that AGOs selectively load endogenous single-stranded (ss)RNAs, suggesting that AGO loading may conform to sequence specificity. To understand ssRNA sequence discrimination by AGOs, we have developed HIgh-throughput Sequencing mediated Specificity Analysis (HISSA), a universal method to study sequence specificity of RNA binding proteins. HISSA allows massively parallel analysis of RNA binding efficiency by using randomized oligos in in-vitro binding assays and quantifying RNAs by deep-sequencing. Our results revealed fly AGO2 loading to be strongly favored by G-rich sequences. Intriguingly, fly AGO1 showed two distinct mechanisms for ssRNA loading. One of which is associated with a 5’ end ‘GAC’ motif in addition to an A/U-rich backbone while another showed a strong preference for a ‘U’ at the 5’ terminus with GC-rich sequences in the seed region. We seek to further explore these two underlying mechanisms in selective ssRNA-AGO1 loading. Our findings may have implications in designing therapeutic ssRNAs for gene silencing to achieve high target-specificity and maximal efficacy.