Automated Covariate Screening Methods

Another automated algorithm is the WAM (Wald s Approximation Method) algorithm developed by Kowalski and Hutmacher (2002). They showed that ranking Schwarz s Bayesian Criterion (larger is better) 

Using actual data sets, Kowalski (2001) showed that five out of six case studies selected the same model as stepwise procedures but did not perform as well when the data were rich and FO-approximation was used. He concluded that WAM might actually perform better than FOCE at choosing a model. However, one potential drawback for this approach is that it requires successful estimation of the variance-covariance matrix, which can sometimes require special handling to develop (e.g., if the model is sensitive to initial estimates, the variance-covariance matrix may not be readily evaluated). Therefore, the WAM algorithm may not be suitable for automated searches if the model output does not always include the standard errors. 

As an example, Phase 1 concentration-time data from 300 subjects (150 males and females two to five observations per subject) were simulated using a 

1-compartment model with first-order absorption of 20 mg Drug X with complete absorption. Clearance was linearly related to weight (in kg) and volume of distribution was 35% higher in females than males as follows 

Covariate on pharmacokinetic parameter NONMEM results WAM results 

---

Covariate screening methods are used when there are a large number of covariates, such that evaluating every possible combination in a model is prohibitive. With this methodology, EBEs of the random effects are treated as data and then exploratory methods are used to assess the relationship between the random effects and the covariate of interest. In other words, each individual s pharmacokinetic parameter, clearance for example, is estimated and treated as a being without measurement error. These Bayes estimates are then compared against subject-specific covariates for a relationship using either manual or automated methods. 

In a subsequent study, the DemixC method was appHed to a mixture of D-glucose, L-histidine, L-lysine, serotonin hydrochloride, D-sorbitol as well as to the venom of the walking stick insect Anisomorpha buprestoides, which consists of at least six compounds [84]. The H—H TOCSY spectra of 2048 X 512 points were recorded in a screening setup at 600 MHz with a 1 mm probehead. With the help of databases, the mixture was deconvolved and the venom identified. Thus, automated mixture deconvolution in screening mode using covariance processed H—H TOCSY experiments was found feasible. 

---