Using Structural and Evolutionary Information to Detect and Correct Pyrosequencing Errors in Noncoding RNAs. - Archive ouverte HAL Access content directly
Journal Articles Journal of Computational Biology Year : 2013

Using Structural and Evolutionary Information to Detect and Correct Pyrosequencing Errors in Noncoding RNAs.

(1) , (2, 3) , (4)
1
2
3
4

Abstract

The analysis of the sequence-structure relationship in RNA molecules is not only essential for evolutionary studies but also for concrete applications such as error-correction in next generation sequencing (NGS) technologies. The prohibitive sizes of the mutational and conformational landscapes, combined with the volume of data to process, require efficient algorithms to compute sequence-structure properties. In this article, we address the correction of NGS errors by calculating which mutations most increase the likelihood of a sequence to a given structure and RNA family. We introduce RNApyro, an efficient, linear time and space inside-outside algorithm that computes exact mutational probabilities under secondary structure and evolutionary constraints given as a multiple sequence alignment with a consensus structure. We develop a scoring scheme combining classical stacking base-pair energies to novel isostericity scores and apply our techniques to correct pointwise errors in 5s and 16s rRNA sequences. Our results suggest that RNApyro is a promising algorithm to complement existing tools in the NGS error-correction pipeline.
Fichier principal
Vignette du fichier
main_RNAPyro_JCB.pdf (706.08 Ko) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-00828062 , version 1 (30-05-2013)

Identifiers

Cite

Vladimir Reinharz, Yann Ponty, Jérôme Waldispühl. Using Structural and Evolutionary Information to Detect and Correct Pyrosequencing Errors in Noncoding RNAs.. Journal of Computational Biology, 2013, 20 (11), pp.905-19. ⟨10.1089/cmb.2013.0085⟩. ⟨hal-00828062⟩
258 View
145 Download

Altmetric

Share

Gmail Facebook Twitter LinkedIn More