Skip to content

Latest commit

 

History

History
96 lines (66 loc) · 2.67 KB

README.md

File metadata and controls

96 lines (66 loc) · 2.67 KB

RNACons

Description

RNACons is an algorithm that finds RNA secondary structure consensus. That is, with a list of sequences and potential structures, it tries to find a group of structures, one per sequence, who have the greatest similarity between all of them.

To do so, it first represents RNA secondary structures as base pair trees. Base pair trees are ordered rooted trees built from the base pairs of a RNA secondary structure. A special tree edit distance is used to compare these trees. Potentially many equivalent consensus are found during this phase.

The second step refines the consensus previously created by minimizing the string edit distance on the Vienna dot-bracket representation of RNA secondary structure fitting the base pair trees. This step insures that unpaired regions match as well as possible.

RNACons versions

Two versions of the consensus optimizer available, one using an exact solver with branch and bound strategy and a heuristic one using an hybrid genetic algorithm. Each solver can consider suboptimal consensus solutions during the first phase to preserve more variety in the final solutions.

Shape Explorer

Shape explorer is a tool to explore RNACons outputs. This is useful mostly when RNACons considers suboptimal solutions. It filters consensus by keeping only the best consensus for each arrangement of abstract shapes level 5.

Requirements

  • Java JRE 7 or greater

Input Files

Input files must be of the following fasta-like format. That is it should look like this (here suboptimal_i_j refers to the jth suboptimal structure of the ith molecule).

All suboptimal structures must be represented in Vienna dot bracket notation.

>name_0
sequence_0
suboptimal_0_0
suboptimal_0_1
...
suboptimal_0_n
>name_1
sequence_1
suboptimal_1_0
suboptimal_1_1
...
suboptimal_1_m
>name_
...

Check out data/examples/mccons_example.marna for a real example.

Output Format

Consensus are outputted in the following fasta-like format to standard output (suboptimal_n_l would be the chosen suboptimal at index l of the nth molecule).

The tree score and string edit score are the average distance between all selected structures.

> solution_index coarseScore refinedScore
suboptimal_0_i
suboptimal_1_j
suboptimal_2_k
...
suboptimal_n_l
> solution_index coarseScore refinedScore
...

Check out data/examples/shape_explorer_example.marna for a real example.

To Do