ZUKER - RNA secondary structure prediction

The ZUKER program predicts the optimal unknotted structure of an RNA sequence based on the model proposed by Tinoco et. al. This model states that the free energy can be computed as a sum of independent energies of the loops of the structure. By the theory of thermodynamics, the optimal structure is the structure of minimal energy.

Installation

The ZUKER program has been implemented in C. To install ZUKER first download the source code, then unzip (gunzip zuker.tar.gz) and dearchive (tar xvf zuker.tar) it. The source code files and data files for energy parametres at 37 °C should now be available in the current directory and the subdirectory energyfiles. The program can be compiled by running make.

Usage

The program is used as

zuker [-v] [file]

which determines the optimal structure for the sequence specified in file; if no filename is given the sequence is taken from the standard input. The '-v' option determines whether the structure is output in the vienna format as well as a list of base pairs and energy increments.

Suggestion

If you simply need to compute some RNA secondary structures, I recommend either using Michael Zuker's mfold server or downloading the Vienna RNA Package, as these come with grafical interfaces and all other kinds of bells and whistles. The only three situations where I would recommend using ZUKER are if you have reason to believe that the predicted structure should contain interior loops with more than thirty unpaired bases (the standard cutoff size), if you really need the extra speed obtained by the improved handling of interior loops (don't expect more than at most a 25% reduction in running time), or if you are looking for a rather concise implementation of the RNA secondary structure prediction algorithm to which you can add your own modifications and extensions (note: I'm gonna be really pissed if you use my code to develop a publicly available RNA secondary structure prediction program that only runs on Windows - otherwise, feel free to cannibalise).

References

The algorithm used for predicting RNA secondary structures is thoroughly described in

Fast evaluation of internal loops in RNA secondary structure prediction, Rune Lyngsų, Michael Zuker and Christian N. S. Pedersen, published in Bioinformatics, vol. 15(6), pp. 440-445.

A shorter description of the algorithm combined with a study of internal loop sizes can be found in

Internal loops in RNA secondary structure prediction, Rune Lyngsų, Michael Zuker and Christian N. S. Pedersen, presented at RECOMB'99.

These two articles have been merged for publication in the BRICS report series in

An improved algorithm for RNA secondary structure prediction, Rune Lyngsų, Michael Zuker and Christian N. S. Pedersen.

The energy model of Tinoco et. al. was proposed in

Estimation of secondary structure in ribonucleic acids, Ignacio Tinoco, Olke C. Uhlenbeck and Mark D. Levine, Nature 230, pp. 362 - 367, 1971,

Improved estimation of secondary structure in ribonucleic acids, Ignacio Tinoco, Philip N. Borer, Barbara Dengler, Mark D. Levine, Olke C. Uhlenbeck, Donald M. Crothers and jay Gralla, Nature New Biology 246, pp. 40 - 41, 1973.


Rune Lyngsų, rlyngsoe@daimi.au.dk