Using a straightforward maximum-likelihood approach for estimating heritability in this model requires knowledge of the identity of the causal SNPs, and hence the covariance matrix.
Most straightforward from maximum likelihood theory is the Unconditional Maximum Likelihood approach (UML; or Joint Maximum Likelihood, JML; cf. Baker and Kim 2004, ch. 5.6).
We preferred a Bayesian approach based on simulation studies that show it to be more straightforward than Maximum Likelihood methods for estimating gene flow using a single locus [ 57] (the nuclear locus was monomorphic for all samples included in the gene flow analysis).
Parameter estimation by maximum likelihood is straightforward.
For each strain i, given the set of probes representing a C58 replicon a on the microarray, it is straightforward to compute a likelihood function on the basis of this modeling.
If transmission dynamics were perfectly observed, it would be straightforward to calculate the likelihood of the data (comprising admission, isolation, and discharge times, as well as screening dates and results) given parameters θ = { p, z, a0, a1, a2}.
Five variants of this readmission model are considered (all with straightforward modifications to the likelihood): In model III, patients who are not colonized at discharge are assumed to have the same phase-independent probability, ν, of being colonized on subsequent readmissions as patients with no previous documented admissions.
Maximum likelihood estimation is straightforward and for a first-order dynamic equation, as in (13), an analytic expression for the information matrix is available.
The position-specific statistics are computationally efficient, straightforward to obtain from maximized likelihood functions and allow a relatively straightforward interpretation of the resulting statistic.
The derivation of the objective function to minimize in order to find the parameters that maximize the likelihood is a straightforward generalization of the single substrate/product case.
Computing the likelihood function is straightforward because once the model parameters have been specified, we have all the equations needed to simulate time courses and calculate likelihoods based on the specified noise model.
