Seminars

Joint Statistics / Biostatistics and Medical Informatics Seminar

High-Dimensional Semilinear Model for Analysis of Microarray Data:
Theoretical Exploration and Methodological Development

Jianqing Fan, PhD,
Department of Operations Research and Financial Engineering, Princeton University

Wednesday, February 25, 2004, 4-5 pm

1221 Computer Sciences and Statistics building, 1210 West Dayton St.

ABSTRACT

Normalization of microarray data is essential for coping with experimental variations and revealing meaningful biological results. We have developed a normalization procedure based on within-array replications via a Semi-Linear In-slide Model (SLIM), which adjusts objectively experimental variations without making critical biological assumptions. This semiparametric model has a number of interesting features: the parametric component and the nonparametric component that are of primary interest can be consistently estimated, the former possessing parametric rate and the latter having nonparametric rate, while the nuisance parameters can not be consistently estimated. This is an interesting extension of the partial consistent phenomena observed by Neyman and Scott (1948). The significant analysis of gene expressions is based on a newly developed weighted t-statistic, which accounts for the heteroscedasticity of the observed log-ratios of expressions, and a balanced sign permutation test. We illustrated the use of the newly developed techniques in a comparison of the expression profiles of neuroblastoma cells that were stimulated with a growth factor, macophage migration inhibitory factor.

If time allows, the second part of the talk will discuss recent advances in multiple testing methods based on False Discovery Control. We treat the fraction of false discoveries (as a function of the rejection threshold) as a latent stochastic process and we find a confidence envelope for this process. From this, we are able to attain strong control of the error rate. The techniques extend in an interesting way to spatial and clustering problems on random fields.

This is joint work with Larry Wasserman.

Return to seminar list