One of the studies was used to define the derivation cohort model in which the authors identified the prediction parameters.
Subsequently, we used logistic regression analysis to identify the prediction factors for clustering of individual patients in the distinct BR patterns.
To identify the prediction of IPA in CIIC patients, we applied multivariate logistic regression analysis for the parameters that demonstrated a P value less than 0.05 in the univariate analysis.
Using the optimal cut-offs identified, the prediction models were more sensitive at picking up proven MODY cases (539/594 [91% sensitivity]), with similar specificity (560/597 [94%]), correctly classifying more patients overall (92% vs 81%, p < 0.0001).
Before conducting our numerical simulations, we analyzed GCM outputs to identify the predictions that would maximize and minimize VC.
The method proposed here is based on identifying the recombinant predictions which are supported by the three methods.
As far as possible the algorithm should be chemically meaningful, however fundamentally the interpretation is identifying the cause behind the prediction and not the cause of activity.
Clearly, by using the modelling module in ChemSAR, one could conveniently construct different algorithm models for one dataset and then makes a comprehensive comparison and further analysis to identify the best prediction model for the current problem.
These data are then used in different combinations to identify the best prediction methodology.
We used logistic modeling to identify the optimal prediction model of group membership.
