Covariate-Adjusted Classification Trees (COVACT) with Applications to Neuroscience

• Activity Number: 470
• Type: Contributed
• Date/Time: Wednesday, August 1, 2012 : 8:30 AM to 10:20 AM
• Sponsor: Biometrics Section
• Author(s): Josephine Asafu-Adjei*+ and Allan Sampson
• Companies: Harvard School of Public Health and University of Pittsburgh
• Keywords: Classification trees ; Covariates ; Post-mortem tissue studies ; Schizophrenia

Abstract

In studies that compare several diagnostic or treatment groups, subjects may not only be measured on a certain set of feature variables of interest, but also measured on additional subject characteristics (covariates). The results from multiple studies done on the same groups of subjects can be integrated using statistical discrimination techniques, such as classification trees, to identify which feature variables best distinguish among the groups under consideration. However, in order to determine the true discriminatory power of the feature data, it is important to account for covariate effects. We present a modified approach to classification trees that accounts for covariate effects on the feature data. The methodology we develop is then applied to a series of post-mortem brain tissue studies conducted by Sweet et al. (2003, 2004, 2007, 2008), with the aim of comparing the neurobiological characteristics of subjects with schizophrenia to those of normal controls.