Rp-Bp: Ribosome profiling with Bayesian predictions#
Introduction#
Ribosome profiling (Ribo-seq) is an RNA-sequencing-based readout of RNA translation. Isolation and deep-sequencing of ribosome-protected RNA fragments (ribosome footprints) provides a genome-wide snapshot of the translatome at sub-codon resolution. Aligned by their P-site positions, the footprints from actively translating ribosomes should exhibit a 3-nt periodicity. To select reads and predict translation, most methods, including Rp-Bp, take advantage of this periodic signal.
Rp-Bp is an unsupervised Bayesian approach to predict translated open reading frames (ORFs) from ribosome profiles, using the automatic Bayesian Periodic fragment length and ribosome P-site offset Selection (BPPS), i.e. read lengths and ribosome P-site offsets are inferred from the data, without supervision. Rp-Bp is able to handle de novo translatome annotation by directly assessing the periodicity of the Ribo-seq signal.
Rp-Bp can be used for ORF discovery, or simply to estimate periodicity in a set of Ribo-seq replicates, e.g. to know which samples and read lengths are usable for downstream analyses. When used for ORF discovery, Rp-Bp automatically classifies ORFs into different biotypes or categories, relative to their host transcript.