A vector of inter-crossover distances in cM. This should include
distances from start of chromosome to first crossover, last crossover to end
of chromosome, and chromosome length, if there are no crossovers.
Alternatively, this may be a matrix with the first column being the
distances and second column being the censoring types (censor).
censor
A vector of the same length as d, indicating the
censoring type for each distance. 0 = uncensored, 1 =
right-censored, 2 = initial crossover on chromosome, 3 = whole
chromosome.
nu
A vector of interference parameters (nu) at which to
calculate the log likelihood. If missing, lo and hi must be
specified.
lo
If nu is unspecified, lo indicates the lower value
of the interval in which to search for the MLE. If supint=TRUE, this
should be below the lower limit of the support interval.
hi
If nu is unspecified, hi indicates the upper value
of the interval in which to search for the MLE. If supint=TRUE, this
should be above the upper limit of the support interval.
se
If TRUE and nu was not specified, an estimated SE (based on
the second derivative of the log likelihood) is estimated.
supint
If TRUE and nu was not specified, a likelihood support
interval is calculated, with drop being the amount to drop in log
(base 10).
rescale
If TRUE and nu was specified, re-scale the log
likelihoods so that the maximum is at 0.
drop
If supint was specified, this indicates the amount to
drop in log (base 10) for the likelihood support interval.
tol
Tolerance for converence to calculate the likelihood, SE, and
likelihood support interval.
maxit
Maximum number of iterations in estimating the SE and
likelihood support interval.
max.conv
Maximum limit for summation in the convolutions to get
inter-crossover distance distribution from the inter-chiasma distance
distributions. This should be greater than the maximum number of chiasmata
on the 4-strand bundle.
integr.tol
Tolerance for convergence of numerical integration.
max.subd
Maximum number of subdivisions in numerical integration.
min.subd
Minimum number of subdivisions in numerical integration.
h
Step used in estimating the second derivative of the log
likelihood.
hstep
factor by which h is decreased in each iteration of the
estimation of the second derivative of the log likelihood.
Details
See Broman and Weber (2000) for details of the method.
We use R's integrate function for numerical integrals,
optimize for optimizing the likelihood, and
uniroot for identifying the endpoints of the likelihood
support interval.
Value
If nu is specified, we return a data frame with two columns:
nu and the corresponding log (base e) likelihood. If
rescale=TRUE, the maximum log likelihood is subtracted off, so that
its maximum is at 0.
If lo and hi is specified, the output contains a single row
with the MLE of nu and the corresponding log likelihood. If
se=TRUE, we also include the estimated SE. If supint=TRUE, we
include two additional rows with the lower and upper limits of the
likelihood support interval.