Grant: Discovery Grants Program - Individual

The goal of this research is to develop novel and rigorous statistical methods to handle complex models that are required to characterize various disease progression processes. The proposed methodology will lead to valuable new insights into many aspects of medical research and will be of great interest to medical and biostatistical researchers. Specific areas to which I intend to contribute are outlined as follows. 1. MULTIVARIATE SURVIVAL DATA: Characterizing associated survival times is typically challenging. Various multiple level structures of survival require different modeling strategies for association structures. An important problem is to assess various model assumptions so that valid inference procedures may be chosen. Standard survival analysis often assumes that covariates appear as a linear form, but this assumption is not adequate enough for many problems. In this research I plan to continue my efforts to explore those challenging problems and develop innovative analysis methods. 2. MICROARRAY DATA: Microarray technology becomes a popular tool for investigation of biological features of a variety of medical problems, and statistical methods for characterizing specific features of microarray studies are in great demand. I intend to devote my efforts to develop powerful statistical methods to address various challenging problems arising from a variety of microarray experiments. 3. MEASUREMENT ERROR: In medical studies it often happens that some collected data are subject to measurement error. When carrying out statistical inference, it is important to account for the effects of mismeasured covariates; otherwise, erroneous or even misleading results may be produced. I intend to develop valid inference methods to account for covariate measurement error for multivariate survival data as well as microarray data.