Related Experiment Videos
STRAL: progressive alignment of non-coding RNA using base pairing probability vectors in quadratic time
Deniz Dalli1, Andreas Wilm, Indra Mainz
1Heinrich-Heine-Universität Düsseldorf, Institut für Physikalische Biologie D-40225 Düsseldorf, Germany.
Bioinformatics (Oxford, England)
|April 15, 2006
Summary
StrAl is a novel heuristic method for aligning structural RNA sequences, improving accuracy by considering base pairing probabilities. This new approach offers better performance than standard methods, especially for divergent sequences.
Area of Science:
- Bioinformatics
- Computational Biology
- Molecular Biology
Background:
- RNA sequence alignment is crucial for applications like phylogeny and structure prediction.
- Aligning structural RNAs (ncRNAs) is challenging due to compensatory mutations that obscure sequence homology.
- Existing methods struggle with multiple structural RNA alignment, despite advances.
Purpose of the Study:
- To develop a robust method for aligning multiple structural RNA sequences.
- To account for RNA structure, specifically base pairing, in the alignment process.
- To improve upon existing RNA alignment tools that do not adequately consider structural information.
Main Methods:
- Introduced StrAl, a heuristic approach reducing sequence-structure alignment to a 2D problem.
- Developed a scoring function incorporating sequence similarity and pairing probabilities.
- Evaluated StrAl's performance against established reference alignments.
Main Results:
- StrAl significantly outperforms standard sequence alignment programs for ncRNA.
- The method shows marked improvement when sequence homology is below 65%.
- StrAl achieves comparable runtime to widely used tools like ClustalW.
Conclusions:
- StrAl provides a more accurate method for aligning structural RNAs.
- The heuristic approach effectively balances sequence and structure information.
- StrAl represents a significant advancement in multiple ncRNA sequence alignment.