Joint modeling of RNase footprint sequencing profiles for genome-wide inference of RNA structure

RNA is far more versatile than was originally thought. First seen as a relatively simple set of molecules involved with protein production, it is now known that there are many other different subgroups of RNAs that perform multiple cellular roles. There are micro (mi)RNAs, which can block or degrade messenger (m)RNA (and hence protein production), small interfering (si)RNAs, which help regulate of gene expression, small nuclear (sn)RNAs that are spliced out of mRNAs, and many more.

To make thing even more complicated, RNA sequences only hint at function. Secondary structures such as loops and single- and double-strand configurations are important components of RNA biology. Determining these structures has been possible one RNA molecule at a time using methods such as x-ray crystallography and nuclear magnetic resonance. But with high-throughput RNA sequencing now common, it’s important to be able to reconstruct RNA secondary structure on a genome-wide scale.

In a paper published online in Nucleic Acids Research on September 22, JAX Assistant Professor Zhengqing Ouyang describes an analysis approach that effectively infers single-strand versus double-strand RNA states—a key element of secondary structure—on a genome-wide scale. Current methodologies infer states by comparing data sets generated by different enzymes, one that cuts double-stranded RNA and the other that cuts only where it’s single-stranded. Unfortunately, there are other factors that confound the data, such as how accessible each RNA nucleotide is to the enzymes. By working with the confounding factors in a novel statistical modeling protocol, Ouyang was able to better infer base pairing states for a joint analysis of RNA data sets from yeast.

Ouyang’s analysis method is able to extract interpretable structural features from nucleotide-level read counts at a genomic scale. The work can be applied to dissecting the functional elements encoded in RNAs and providing a deeper understanding of global RNA regulation.

Nucl. Acids Res. (2015) doi: 10.1093/nar/gkv950 (published online 9/22/15)