A method for detecting the conserved helices of functional RNA structures, including transient, pseudo-knotted and alternative structure
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Failed to load latest commit information.


Readme file for Transat - a program for predicting evolutionarily conserved
helices and their p-values for an input multiple sequence alignments and an
input tree relating the sequences of the input alignment.


The Transat package is copyrighted by Nicholas J. P. Wiebe (2008-2010) and
Irmtraud M. Meyer (2008-2010), see also files LICENSE and COPYRIGHT for more


If you use this package or parts of it, please cite:

* Nicholas J. P. Wiebe and I.M.Meyer, Transat - a method for detecting the conserved helices of functional
  RNA structures, including transient, pseudo-knotted and alternative
  structures, PLoS Computational Biology (2010), 6(6): e1000823

How to compile and run Transat:

You will need to compile your own Transat executable as some features of
Transat will be specific to your own computer and file system.

(1) Transat requires the alignment shuffling perl script from RNAz. RNAz can be found at http://www.tbi.univie.ac.at/~wash/RNAz/
    The script is called rnazRandomizeAln.pl. Download it.

(2) Specify the location of the perl-script rnazRandomizeAln.pl in the makefile for Transat which is source/Transat/makefile
    and set the value the RNAZ_SHUFFLER_LOC variable in the makefile to the appropriate location in your filesystem.  
    We suggest you do this by specifying its entire path rather than by giving a relative location to the makefile. If you
    plan to run Transat jobs on a different computer than the one used for compiling the Transat executable, make sure you choose a location 
    for rnazRandomizeAln.pl which you can access when running jobs elsewhere.

(3) Compile your own Transat executable by

    - going to directory source/Transat/ 
    - saying "make clean"
    - saying "make all" which should result in an executable file called "Transat" being produced. This is the file for running
      Transat analyses. 

(4) optional: Install T-Coffee, if you want Transat to realign an input alignment before shuffling, and ensure that the command "t_coffee" is on your PATH.
    Type "which t_coffee" in a shell to check this is the case.

(5) Running Transat jobs:

    two types of command line structure:

    Transat -fasta <fasta alignment file> -ct <consensus structure .ct file> -tree <tree in newick format>
    Transat -fasta <fasta alignment file, with last entry as the structure in dot-bracket notation> -tree <tree in newick format>


    -fasta [filename] : file containing alignment, in aligned fasta format, with the last entry as the structure in dot-bracket notation (with allowed brackets '()', '[]', '<>', "Aa", and "Bb") unless the option -ct is used
    -ct [filename] : ct-like input file containing known structure. If a ct file not specified, then Transat will assume that the last line of the alignment file contains 
     the structure in dot-bracket notation.
    -tree [filename] : file containing phylogenetic tree in newick format.
    -minSL [int] : sets the minimum helix length. Only competing helices with a length greater than this value will be stored (default: 8, but the Transat paper uses 3)
    -randomize [int] : sets the number of shuffled alignments are used in the null distributions (default = 500)
    -realign : if this flag is present, Transat does a realignment step before producing a shuffled alignments. Uses t_coffee for realignment step (t_coffee must be on PATH!)
    -noDB : if this flag is present, Transat will not print the dot-bracket structure of each helix to the output file.
    -noStruct : use this flag if you would like to run Transat without specifying a known structure for the alignment. -ct overrides this flag.

    See the sub-directory "examples" for examples on how to run Transat jobs.

(6) Analyzing the Transat output:

    - the output file of Transat is a data table with an entry line for every predicted helix in the alignment
    - all helices in that table are unique
    - the header lines of the different columns in that table are relatively self-explanatory. As you will see, some columns contain information on the predicted helix that is
      relative to whatever input RNA structure was specified up-front (but Transat does not require or use any structural features known up-front to generate its predictions)
    - the p-values are calculated by shuffling the entire alignment and finding all helices as described in the Transat paper.

More information on Transat can be found 

- on the web-page people.cs.ubc.ca/~irmtraud/transat


- irmtraud.meyer@cantab.net


The sub-directory "examples" contains several examples as well as a readme
file explaing how to run Transat jobs for these examples.


November 2009 - release of Transat version 1.0