Bioinformatics: Selecting against accidental RNA interactions

  1. Meredith Corley  Is a corresponding author
  2. Alain Laederach  Is a corresponding author
  1. University of North Carolina, United States

Translation is the process by which the genetic information in a molecule of messenger RNA (mRNA) produces a protein, and the rate at which protein is produced from a given mRNA molecule is called the translation efficiency. This number is different for different mRNA molecules (Maier et al., 2009; Guo et al., 2008), which is why researchers are trying to determine which features of these molecules affect their translation efficiency (Tuller et al., 2010; Ferreira et al., 2013; Kozak, 2005; Gingold and Pilpel, 2011).

Now, in eLife, Paul Gardner of the University of Canterbury and colleagues – including Sinan Umu (as first author), Anthony Poole and Renwick Dobson – report that the translation efficiency in bacteria and archaea is influenced by a phenomenon called "avoidance" (Umu et al., 2016). Avoidance is the degree to which an mRNA molecule avoids random interactions with noncoding RNA molecules in the cell. Noncoding RNAs, as their name suggests, do not code for proteins, but they still make up a majority of the RNA in any given cell. Indeed, the researchers show that the levels of noncoding RNAs in bacterial cells are two orders of magnitude greater than the levels of mRNAs.

To estimate the probability of random base-pairing interactions taking place between mRNAs and noncoding RNAs, consider a five-base region in a single mRNA. This region can have any one of a possible 45=1024 sequences. If the total number of bases from all the noncoding RNAs in the cell is S, then the number of noncoding RNAs in the cell that have a perfectly complementary five-base region is approximately S/1024. Umu et al. studied 325 noncoding RNAs so, assuming an average length of 200 bases for these, we have S ≈ 325*200 ≈ 65000. Therefore, on average, the number of these noncoding RNAs that have a five-base region that is perfectly complementary to the five-base region in the mRNA will be 65000/1024 ≈ 63. Given the number of mRNAs and noncoding RNAs that are found in cells, random interactions between the two are inevitable. However, if we find that a given mRNA has base-pairing interactions with fewer noncoding RNAs than expected, then this is avoidance (Figure 1A).

Random interactions between mRNAs and noncoding RNAs.

(A) A given region of mRNA (pink) should undergo random base-pairing interactions with a certain number of noncoding RNAs (orange). However, a phenomenon called "avoidance" results in the number of observed pairings (right) being fewer than the number expected by chance (left). This simplified picture does not allow for the influence of binding energies and other effects. (B) The ability of RNA molecules to fold into complex structures means that mRNA-noncoding RNA interactions can have a range of binding energies (see color bar). "Avoidance" results in the observed binding energies (right) being, on average, weaker than the binding energies expected by chance (left). Umu et al. report a significant difference in binding energy distributions in 73% of bacterial and archeal species. (C) If a random base-pairing interaction results in a noncoding RNA pairing with part or all of a start codon in an mRNA, the ribosome will not be able to translate the mRNA.

Realistically, RNA binding interactions are governed by thermodynamics and do not always follow strict pairing rules. Furthermore, RNA molecules can pair with themselves in intra-molecular interactions and must “unfold” a given region in order to pair with another molecule. Thus it is important to quantify mRNA-noncoding RNA interactions with net "binding energy" calculations. The binding energy quantifies the thermodynamics of RNA base-pairing, with low binding energies indicating very stable interactions. To explore the phenomenon of avoidance, Umu et al. used a computational RNA interaction model to estimate the binding energies for interactions between mRNAs and noncoding RNAs. The RNAs include a core set of 114 mRNAs that are well conserved across bacteria (including 40 that are also conserved across archaea) and 325 noncoding RNAs from six families of RNA that are also well conserved across bacterial and archaeal species.

They found that, on average, the interactions between the core noncoding RNAs and mRNAs were weaker than the interactions between the noncoding RNAs and a control set. In other words, they found that the average mRNA "avoids" interactions due to less stable pairing with noncoding RNAs (Figure 1B). This trend holds true for over 70% of the bacteria and archaea that they tested. There are, of course, noncoding RNAs whose primary function is to bind to mRNAs, but these were excluded from the study. Instead, the goal was to observe selection against accidental interactions between mRNAs and the large and diverse set of noncoding RNAs that are resident in the cell.

Umu et al. hypothesize that avoidance is due to the negative effect that the interactions between mRNAs and noncoding RNA could have on translation efficiency: for example, if a noncoding RNA pairs with a start codon in an mRNA, it will prevent translation from taking place because the ribosome will not be able to bind to that mRNA (Figure 1C). To test this hypothesis, the researchers designed and synthesized a set of mRNAs with sequences that have high levels of avoidance, and a set of mRNAs with low levels of avoidance. When they measured the translation efficiency for both sets of mRNAs, they found that it was much higher for the highly-avoidant set.

Umu et al. also synthesized different sets of mRNAs to explore two other factors that are thought to influence translation efficiency: codon bias and the intra-mRNA folding energy (Kudla et al., 2009; Tuller et al., 2010). Both factors did cause some variation in the production of protein, but avoidance was responsible for the most variation. They also found the same correlation with avoidance when they studied previously published measurements of bacterial translation efficiency. This suggests that the ability of an mRNA to avoid interactions with noncoding RNAs is a hitherto unknown, yet important factor affecting translation efficiency.

One notable aspect of this study is that it relied almost entirely on publicly available data sets. This underscores the importance of open data for exploring basic biological questions that apply to many different organisms. Using this data, which no single lab could have collected alone, Umu et al. have shown that mRNA sequences are optimized to minimize interactions with noncoding RNAs and have demonstrated why such avoidance is so desirable. And the need to avoid spurious interactions is not unique to RNA: the emergence of complex life depends on optimizing molecular interactions that lead to reproduction in the midst of molecular chaos. Although networks of highly specific molecular interactions are a hallmark of evolution, in many cases it is just as important to avoid accidental interactions.


Article and author information

Author details

  1. Meredith Corley

    Department of Biology, University of North Carolina, Chapel Hill, United States
    For correspondence
    Competing interests
    The authors declare that no competing interests exist.
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-4485-5663
  2. Alain Laederach

    Department of Biology, University of North Carolina, Chapel Hill, United States
    For correspondence
    Competing interests
    The authors declare that no competing interests exist.
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-5088-9907

Publication history

  1. Version of Record published: September 20, 2016 (version 1)
  2. Version of Record updated: October 3, 2016 (version 2)


© 2016, Corley et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.


  • 1,502
  • 218
  • 1

Views, downloads and citations are aggregated across all versions of this paper published by eLife.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Meredith Corley
  2. Alain Laederach
Bioinformatics: Selecting against accidental RNA interactions
eLife 5:e20686.

Further reading

    1. Computational and Systems Biology
    Natalie R Davidson, Casey S Greene
    Research Article

    Science journalism is a critical way for the public to learn about and benefit from scientific findings. Such journalism shapes the public’s view of the current state of science and legitimizes experts. Journalists can only cite and quote a limited number of sources, who they may discover in their research, including recommendations by other scientists. Biases in either process may influence who is identified and ultimately included as a source. To examine potential biases in science journalism, we analyzed 22,001 non-research articles published by Nature and compared these with Nature-published research articles with respect to predicted gender and name origin. We extracted cited authors’ names and those of quoted speakers. While citations and quotations within a piece do not reflect the entire information-gathering process, they can provide insight into the demographics of visible sources. We then predicted gender and name origin of the cited authors and speakers. We compared articles with a comparator set made up of first and last authors within primary research articles in Nature and a subset of Springer Nature articles in the same time period. In our analysis, we found a skew toward quoting men in Nature science journalism. However, quotation is trending toward equal representation at a faster rate than authorship rates in academic publishing. Gender disparity in Nature quotes was dependent on the article type. We found a significant over-representation of names with predicted Celtic/English origin and under-representation of names with a predicted East Asian origin in both in extracted quotes and journal citations but dampened in citations.

    1. Computational and Systems Biology
    2. Genetics and Genomics
    Chananchida Sang-aram, Robin Browaeys ... Yvan Saeys
    Research Article

    Spatial transcriptomics (ST) technologies allow the profiling of the transcriptome of cells while keeping their spatial context. Since most commercial untargeted ST technologies do not yet operate at single-cell resolution, computational methods such as deconvolution are often used to infer the cell type composition of each sequenced spot. We benchmarked 11 deconvolution methods using 63 silver standards, 3 gold standards, and 2 case studies on liver and melanoma tissues. We developed a simulation engine called synthspot to generate silver standards from single-cell RNA-sequencing data, while gold standards are generated by pooling single cells from targeted ST data. We evaluated methods based on their performance, stability across different reference datasets, and scalability. We found that cell2location and RCTD are the top-performing methods, but surprisingly, a simple regression model outperforms almost half of the dedicated spatial deconvolution methods. Furthermore, we observe that the performance of all methods significantly decreased in datasets with highly abundant or rare cell types. Our results are reproducible in a Nextflow pipeline, which also allows users to generate synthetic data, run deconvolution methods and optionally benchmark them on their dataset (