27

c

c

https://rfam.xfam.org/family/RF00037

Those who want to read more in-depth information about techniques and RNA func­

tions in context can check out our books on regulatory RNA at Google-Books (Dandekar

and Bengert 2002; Dandekar and Sharma 1998).

RNAAnalyzer: A Quick Analysis for Each RNA Molecule

c

c

https://rnaanalyzer.bioapps.biozentrum.uni-­wuerzburg.de

Another way to understand RNA and regulatory elements is to analyze the secondary

structure and sequence motifs through a program. In our program developed for this pur­

pose, the RNAAnalyzer, you can enter any RNA sequence, which is then searched for

regulatory elements. The result is a list of regulatory element hits and important further

descriptions, such as whether there is a lot of secondary structure, whether proteins can

bind to the RNA or whether the RNA molecule is perhaps an mRNA, but also numerous

other pieces of information (Bengert and Dandekar 2003).

One way to further check or supplement these results is to use the AnDom software (cf.

Chap. 1, Protein analyses). For regulatory RNA, another alternative is the RegRNA server

from Taiwan (https://regrna2.mbc.nctu.edu.tw/), which also offers a rapid analysis for

RNA using related methods independently.

RNAfold and mFold Show RNA Structure

Another important method to analyze the RNA structure is to check the RNA folding with

the pairing scheme: A always pairs with U (two hydrogen bonds), G with C (three hydro­

gen bonds). With the help of these rules and other rules (G pairs with U, only one hydrogen

bond; thermodynamic parameters such as the Tinocco parameters), it is possible to sys­

tematically try out with the computer which structural folding of the RNA will lead to the

highest number of base pairings and, in particular, hydrogen bonds and energy. This is also

known as dynamic programming (Eddy 2004), because the sequence is broken down into

small substrings and the optimal RNA structure is calculated iteratively (for longer RNA

molecules, more and more memory is allocated dynamically for the base pairings).

Simple approaches such as the Nussinov algorithm are based on the optimal base pair­

ing of the RNA, whereas extensions additionally consider the folding energy. The best

known is the prediction algorithm of Zuker and Stiegler (1981), e.g. mFold server (https://

unafold.rna.albany.edu/?q=mfold; out of operation since November 1, 2020.) or its further

developments such as the RNAfold server (https://rna.tbi.univie.ac.at/cgi-­bin/

RNAWebSuite/RNAfold.cgi). The Sankoff algorithm takes phylogeny into account in

addition to alignment and folding energy (e.g. LocARNA tool; https://www.bioinf.uni-­

freiburg.de/Software/LocARNA/). However, other software for RNA folding is also avail­

able (e.g., ViennaRNA package; https://www.tbi.univie.ac.at/RNA/; Freiburg RNA tools;

https://rna.informatik.uni-­freiburg.de/). By looking at several folding types (i.e., still the

2.2  Analysis of RNA Sequence, Structure and Function