MODENA website

 

 

THIS IS THE WEBSITE FOR OLD VERSIONS. THE LATEST VERSION IS HERE !

MODENA (Multi-Objective DEsign of Nucleic Acids) is an RNA inverse folding program based on multi-objective genetic algorithm. Non profit, academic users can download and use MODENA. For commercial use, please contact the author. These software are distributed as "no warranty" software. Copyright(C) 2010, Akito Taneda Laboratory.

References

 

Akito Taneda, MODENA: a multi-objective RNA inverse folding, Adv. Appl. Bioinform. Chem. 4: 1-12 (2011). (pubmed)

Akito Taneda, Multi-objective genetic algorithm for pseudoknotted RNA sequence design, Front. Genet. 3:36 (2012). (pubmed) (supplementary material) We updated this supplementary material to correct the order of the MODENA+Hotknots results on 2015 April.

Akito Taneda, A Multi-Objective Evolutionary Algorithm for RNASequence Design, Computational Biology Research Center Workchop 2010, CBRC, Odaiba, Jul. 28-30, 2010 (Poster No. P55).



Requirements

 

1. Vienna RNA package (for RNAfold)

If you would like to perform the RNA inverse folding with CentroidFold (as a direct problem solver), please install

2. CentroidFold

We tested using "Vienna RNA package ver. 1.8.3 on a Linux (Fedora 11, Intel Core2quad x86-64)" and " Vienna RNA package ver. 1.8.3 on a Linux (CentOS 5.3, Intel Core2duo x86-64) .

We used CentroidFold ver 0.0.8.

If you would like to design pseudoknotted RNAs, the following software is required.

3. IPknot

If you use these software for your research, plese cite the papers describing these software. References for these software can be found in the original papers of MODENA.



Installation

 

1. download MODENA (an executable file & test target structure files are included).

Linux (x86_64)
Linux (i386)
Windows
modena.exe
Mac
modena

We tested these files on CentOS 5.3 (Intel Core2duo) and Fedora Core release 11 (CPU: , Intel Core2quad). We believe that users can run MODENA on other RedHat-based Linux platforms.

2. copy the executable file (modena) to the command search path, e.g. /usr/local/bin/.

Currently, I would like to distribute executable files alone.

2011-07-14 : The dataset used in the benchmark can be downloaded from here.

2011-10-17 : We updated the modena.tar.gz for pseudoknot design (v0039). The previous version (v0029) of modena.tar.gz can be downloaded from here.

2012-9-29 : We updated the modena.tar.gz for a bug fix and the i386 architecture (v0040). The previous version (v0039) of modena.tar.gz can be downloaded from here.

2015-01-22 : The dataset used in the pseudoknot design benchmark can be downloaded from here.



Usage

 

1. To design the RNA sequence with a user-given target structure, please type:

modena -f target.in

As a default, RNAfold is used as a direct problem solver. This command will output a number of successfully designed RNA sequences denote by "Individual .... 1.0". They are weak Pareto optimal solutions and have a "Rk = 1". The other RNA sequences with a Rk = 2, 3, 4, ... and so on are dominated solutions i.e. they are designed unsuccessfully. It is noted that even the solutions with a Rk = 1 can have a similarity score < 1.0 (this represents an unsuccessful design). An example of the outputted RNAsequence is shown below:

Individual= 17 Rk= 1 Sc= 7.900000 1.000000

AAAAACACCAGGACAUACCUGGAGAUAG
.......(((((.....)))))......

This example has a structure stability score score epsilon = 7.9 and structure similarity score sigma = 1.0. This is an example of a successful design.

Individual= 9 Rk= 2 Sc= 5.500000 0.928571

CGGAAAAGUGAUCACGGAUCACCGAAAG
.......((((((...))))))......

This example has a structure stability score score epsilon = 5.5 and structure similarity score sigma = 0.93. This is an example of an UNSUCCESSFUL design since similarity score is 0.93, which is not equal to 1.0.

2. To design RNA sequences with CentroidFold (as a direct problem solver), please type:

modena -centf -f target.in

To design pseudoknotted RNA sequences with IPknot (as a direct problem solver), please type:

modena -ipknot -f pseudoknot.in

3. The GA population size can be changed by a "-pop " option, and the maximum GA iteration number can be changed with a "-it" option:

modena -f target.in -pop 100 -it 100

We recommend to set a same number to the population size and iteration number.

To change the initial random number of MODENA, please use a -r option (default: -r 12345)

modena -f target.in -r 12345

To change the crossover parameter n_c, please use a -cross option (default: -cross 1),

where from '-cross 1' to '-cross 5' is available.

modena -f target.in -cross 2

4. For pseudoknot specification in the target structure file,, we can use ( ), < >, [ ] , { } , A a, B b, C c, D d, and E e to denote a base pair. For example, please see pseudoknot.in file in the modena????.tar.gz.


Comments, bug reports are welcome. E-mail: taneda (at) cc.hirosaki-u.ac.jp