Parseq – an RNA-Seq read count emission model for transcriptional landscape reconstruction with state-space models

Parseq is a statistical approach for transcription landscape reconstruction at a basepair resolution from RNA Seq read counts.

It is based on a state-space model which describes, in terms of abrupt shifts and more progressive drifts, the transcription level dynamics along the genome. Alongside variations of transcription level, Parseq incorporates a component of short-range variation to pull apart local artifacts causing correlated dispersion. Reconstruction of the transcription level relies on a conditional sequential Monte Carlo approach that is combined with parameter estimation in a Markov chain Monte Carlo algorithm known as particle Gibbs. The method allows to estimate the local transcription level, to call transcribed regions, and to identify the transcript borders.

Availability – Parseq is available at: http://www.lgm.upmc.fr/parseq/

Reference