transitionProbability {VanillaICE} | R Documentation |
Wrapper for computing the locus-specific transition probability
transitionProbability(chromosome, position, TAUP = 1e+08, chromosomeAnnotation, verbose = FALSE)
chromosome |
chromosome (integer representation) |
position |
physical position |
TAUP |
Scalar for computing transition probabilities (see Details). |
chromosomeAnnotation |
Optional: chromosome annotation |
verbose |
Logical: verbose output |
The HMM uses locus-specific transition probabilities that are calculated as a function of the physical distance between loci. Specifically, the probability that the locus at position t-1 is not informative for the locus at position t is calculated as 1-exp(-d/TAUP), where d is the physical distance between locus t and locus t-1. The default for TAUP is 1 x 10^8 and can be specified to acheive a desired amount of sensitivity and specificity. Larger values of TAUP decreases the probability of transitioning to other states, and therefore provides a more smooth fit.
The transitionProbability function (i) transforms the physical distance between adjacent loci to an estimate of the genomic distance and (ii) adds an 'arm' variable to the annotation matrix.
chromosome |
chromosome |
position |
physical position |
arm |
an integer. The HMM uses the arm variable as a factor and is fit independently to each 'arm'. |
transitionPr |
locus-specific transition probabilities |
R. Scharpf