Skip to content
Chaining sequence/structure seeds for computing RNA similarity
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.

                   RNA-unchained v1.0 Overview

RNA-unchained (Chaining sequence/structure seeds for computing RNA similarity) aims at detecting structural homologs of a given query RNA sequence in a, potentially large, collection of RNA structures.

Details of the algorithm can be found at

RNA-unchained consists of three steps:

(1) The computation of an index of RNA which contains for each RNA all identified seeds. This index can deal with multiple seeds size and can be modified to add new RNA. (2) Given a file containing one or multiple query it finds all common seeds between query and RNAs of the index and compute the corresponding anchor (chain of seeds). (3) Given the chain, a constrained alignment is performed by the Locarna software. The alignment is computed only if there is t least one anchor between the two considered RNAs.

Version: 1.0 Contact: Laetitia Bourgeade, Cedric Chauve,


RNA-unchained runs with python v2.7 or higher but not python v3.0 or higher. No other compilation is necessary, except installing and configuring other packages.

The following software packages are required

-- LocARNA v1.7.5 or higher

-- ViennaRNA v2.1.1 or higher

If they are not installed to '/usr/local',which is typically the default directory when installed as root, you have to add the path in the shell path.

You can use directly RNA-unchained by adding the path of RNA-unchained folder in the shell path.


Chaining in sequence and structure RNAs before aligning them:

(1) Input file must be at fasta format. It can contains struture but if not RNAfold will be used to compute Minimum Free Enerdy (MFE) struture. (2) Given a fasta file, an index contening all seeds of given sizes is constituted. (3) Given a fasta file, all pairwise analysis between each RNA of the file and each RNA of the index database is done. It generates a file per analysis containing the constraints. (4) Two files one with alignments and one with alignments score is generated with Locarna using the constraints file computed before.

Run RNA-unchained with the command

Command: [options] file.fa

Calculate pairwise alignment using chaining and seeds method

positional arguments: file.fa Fasta file containing RNAs

optional arguments: -h --help show this help message and exit -c, --chaining use the chaining algorithm to find best chain with one type of seeds -a, --alignment align with constraints -m, --makeIndex Construct Index with name DBNAME --add add new RNA to DB -db DBNAME, --dbName DBNAME give the name DBNAME to the index database -r, --r2 chaining only r2 seeds Not compatible with rfb option -rfb, --r2fb seeds optimization. Not compatible with r option -epc, --epcLCS4l anchor optimisation -f forced alignment with Locarna even if there is no constraints -ld l d, --ldSeeds l d seq length and str length for seed --version show program's version number and exit


RNA-unchained will generate the following files in the folder with the name of the index db (if no name given a random one is used)

(1) An 'RNAunchainedProjects' folder is created in your home. It will contains all RNA-unchained analysis. (2) A folder with the specified or random name of the index is created in 'RNAunchainedProjects'. It will contains all analysis computed with this database. (3) The 'indexDB' folder contains the index database. (4) The 'results' folder contains all analysis with one folder with the name of the RNA per query RNA. (5) For each RNA query there is 3 files: - RNAname_constraints: contains computed constraints - chaining_alignments: contains alignments - chaining_score: contains scores of alignments

You can’t perform that action at this time.