GSLCCA estimates the parameters of a given nonlinear model to maximize the correlation with a linear combination of multiple response variables.
GSLCCA was developed to characterize brain activity monitored using EEG (electroencephalogram) under different treatment regimens in research and development projects at Pfizer.
EEG results are typically summarised as a matrix of power spectra, giving the power (microvolts squared) over multiple frequencies (Hz) recorded at different times. GSLCCA is then used to find a linear combination of the powers that is maximally correlated to a nonlinear pharmacometric model.
gslcca package can be used to relate any multivariate
response to any function of time, however it provides shortcuts for
specifying standard PK/PD models (Critical Exponential or Double
Exponential). It also has options to analyse subjects separately;
to estimate parameters separately for different treatments and to
partial out covariates prior to analysis.
Utility functions are provided to plot the results and to investigating the effect of pre-smoothing on the analysis.
Foteini Strimenopoulou and Heather Turner
Maintainer: Heather Turner <firstname.lastname@example.org>
Brain, P., Strimenopoulou, F. and Ivarsson, M. (2011). Analysing electroencephalogram (EEG) data using Extended Semi-Linear Canonical Correlation Analysis. Submitted.
gslcca for the main function, with links to associated functions.