Applying Hidden Markov Models to RNA-seq data

• Enterococcus faecalis is one of the most controversial commensal bacteria of the human intestinal flora that is also responsible for lethal nosocomial infections. Determining the factors that influence its pathogenicity is at present a great challenge. Cutting-edge approaches analyze the E. faecalis bacterium trough the next generation RNA-sequencing technology. Since next generation sequencing is recent and yields a large amount of data, there is a continuous need for appropriate statistical methods to interpret its output. We propose an approach based on hidden Markov models to explore RNA-seq data and show an example of how we can apply this statistical tool to detect transcription start sites. We compare this application with a previously developed method based on signal processing.

Ämnesord

NATURVETENSKAP  -- Data- och informationsvetenskap -- Bioinformatik (hsv//swe)

NATURAL SCIENCES  -- Computer and Information Sciences -- Bioinformatics (hsv//eng)

Nyckelord

RNA-sequencing

statistical modeling

hidden Markov models

inference

Baum-Welch algorithm

auto-regressive hidden Markov model

transcription start sites prediction

Publikations- och innehållstyp

vet (ämneskategori)

rap (ämneskategori)