computeH: computeH

In shimlab/riboHMM2: Analysis tool for Ribo-seq

Description

computeH calculates H_kj (t) = P( Z[t]=k, Z[t+1]=j | x[1:n] )

Usage

computeH(x, RNA, trans, alpha, beta, E, la, lb)

Arguments

x	a vector of observed states
RNA	a 0-1 vector. 1 if next 3-base is stop codon
trans	a vector c(rho_u, rho, delta)
alpha	shape parameter in gamma distribution
beta	rate parameter in gamma distribution
E	a scalar. Normalizing constant for the observed chain x.
la	a matrix from Forward Algorithm
lb	a matrix from Backward Algorithm

Value

A matrix H

Examples

df <- uORF[[1]]
x=df$x; RNA=df$RNA; trans=df$trans; a=df$v; b=df$v/df$m; E=df$E

la <- forwardAlg(x, RNA, trans, a, b, E)
lb <- backwardAlg(x, RNA, trans, a, b, E)
L <- computeL(la, lb)
H <- computeH(x, RNA, trans, a, b, E, la, lb)

H <- round(H, 2)
View(H)
df$z

shimlab/riboHMM2 documentation built on May 19, 2019, 6:23 p.m.