10.6: Evolution of RNA

Last updated
Save as PDF

Page ID: 40979

\( \newcommand{\vecs}[1]{\overset { \scriptstyle \rightharpoonup} {\mathbf{#1}} } \) \( \newcommand{\vecd}[1]{\overset{-\!-\!\rightharpoonup}{\vphantom{a}\smash {#1}}} \)\(\newcommand{\id}{\mathrm{id}}\) \( \newcommand{\Span}{\mathrm{span}}\) \( \newcommand{\kernel}{\mathrm{null}\,}\) \( \newcommand{\range}{\mathrm{range}\,}\) \( \newcommand{\RealPart}{\mathrm{Re}}\) \( \newcommand{\ImaginaryPart}{\mathrm{Im}}\) \( \newcommand{\Argument}{\mathrm{Arg}}\) \( \newcommand{\norm}[1]{\| #1 \|}\) \( \newcommand{\inner}[2]{\langle #1, #2 \rangle}\) \( \newcommand{\Span}{\mathrm{span}}\) \(\newcommand{\id}{\mathrm{id}}\) \( \newcommand{\Span}{\mathrm{span}}\) \( \newcommand{\kernel}{\mathrm{null}\,}\) \( \newcommand{\range}{\mathrm{range}\,}\) \( \newcommand{\RealPart}{\mathrm{Re}}\) \( \newcommand{\ImaginaryPart}{\mathrm{Im}}\) \( \newcommand{\Argument}{\mathrm{Arg}}\) \( \newcommand{\norm}[1]{\| #1 \|}\) \( \newcommand{\inner}[2]{\langle #1, #2 \rangle}\) \( \newcommand{\Span}{\mathrm{span}}\)\(\newcommand{\AA}{\unicode[.8,0]{x212B}}\)

It is useful to understand the evolution of RNA structure, because it unveils valuable data, and can also give us hints to refine our structure predictions. When we look into functionally important RNAs over time, we realize their nucleotides have changed at some parts, but their structure is well-conserved.

In RNA there are a lot of compensatory mutations and consistent mutations. In a consistent mutation, the structure doesnt change e.g. an AU pair mutates to form a G pair. In a compensatory mutation there are actually two mutations, one disrupts the structure, but the second mutation restores it, for example an AU pair changes to a CU which does not pair well, but in turn the U mutates to a G to restore a CG pair. In an ideal world, if we have this knowledge, this is the be the key to predict the RNA structure, because evolution never lies. We can calculate the mutual information content for two different RNAs and compare it. In other words, you compare the probabilities of two base pair structures agreeing randomly vs. if they have evolved to be conserve the structure.

The mutual information content is calculated via this formula:

\[M_{i j}=\sum_{X, Y} f_{i j}(X Y) \log \frac{f_{i j}(X Y)}{f_{i}(X) f_{j}(Y)} \nonumber \]

If we normalize these probabilities, and store the MI in bits, we can plot it in a 3D model and track the evolutionary signatures. In fact, this was the method for determining the structure of ribosomal RNAs long before they were found by crystallography.

The real problem is that we dont have so much information, so what we usually do is combine the folding prediction methods with phylogenetic information in order to get a reliable prediction. The most common way to do this is to combine to Zuker algorithm with some covariance scores. For example, we add stabilizing energy if we have a compensatory mutation, and destabilizing energy if we have a single nucleotide mutation.