|
Abstract:
|
Over the last decade, innovations in genomic technology have led to an explosion of biomarkers discovery for medical diagnosis and drug development. Biomarker discovery entails the identification of putative markers for the response of interest, typically from exploratory focused multi-array and other technology platforms. This is followed by the process of qualification of these biomarkers via additional studies/cohorts to demonstrate their predictive ability for the intended purpose. To successfully meet these goals, statistical techniques necessary for analyzing experimental results are still being developed, while the specialty of genomic experiments continues to introduce many challenging statistical topics. This presentation will detail some of these statistical challenges, including data pre-processing, normalization, gene expression index, calibration, and clinical validation.
|
