The discrepancies between parametric and nonparametric simulations are discussed in light of our dataset and known attributes of both approaches.
Taking these observations into account, all genes of the CODH-hydrogenase gene cluster were investigated for HGT by means of detection of phylogenetic discrepancies and parametric methods.
The parametric bootstrapping simulations showed that this discrepancy between the LRTs was not due to biases in the chi-square approximation to the likelihood ratio statistic.
Here, we present maximum mean methylation discrepancy (MD), a non-parametric statistical test for identifying DMRs from pre-defined regions, explicitly accounting for shape changes in methylation profiles.
The second method is without dependence on known transcript structure, and uses a non-parametric kernel-based statistical test called maximum mean discrepancy.
This is consistent with an earlier observation of the discrepancy of rate estimation by the simple parametric model vs. the complicated rate model and maximum likelihood method at the high GC% isochores [ 38].
The general discussion is applied and examined for special kernel function, discrepancy measures for matrices on Pk,m−k and starting parametric model.
To further evaluate the discrepancies between the different partitions we performed a more parametric phylogenetic approach, Bayesian phylogenetic analysis, using the MKV model [ 42] as implemented in MrBayes v2.2 [ 43], and an extending encoding of the CNs.
In line with Coricelli et al.'s [12] procedure, in a second analysis we modeled a linear parametric modulation by the degree of satisfaction/disappointment, i.e., the amount of discrepancy between the obtained and unobtained outcomes in the chosen gamble only.
The nonlinear parametric robot model is linearized in the same positions and the optimal parameters are obtained by minimizing the discrepancy between the nonparametric FRFs and the parametric FRFs (the FRFs of the linearized parametric robot model).
