Software for(Last Update: March 22nd, 2007. By Shuangge Ma)
Regularized Microarray Meta-Analysis
Shuangge Ma, Ph.D.
Department of Epidemiology and Public Health
Yale University, New Haven, CT 06520
Jian Huang, Ph.D.
Department of Statistics & Actuarial Science and Program in Public Health Genetics
University of Iowa, Iowa, IA 52242
We provide the source code written in R for regularized estimation and biomarker selection in microarray meta analysis using the Meta Threshold Gradient Descent Regularization (MTGDR) method proposed in the manuscript coauthored by Shuangge Ma and Jian Huang. Below is the abstract of the paper. We strongly recommand you reading the full paper before using the software.
In pharmacogenetic studies, it is common that multiple microarray studies are conducted to investigate the relationship between a phenotype and gene expressions. An important goal of such studies is to discover influential genes that can be used as disease biomarkers and construct predictive models. To increase statistical power, meta analysis should be used to combine results from these studies. However, it is difficult to apply the standard meta analysis approaches because of high-dimensionality of microarray data and because different microarray platforms and experimental settings used in different studies may not be directly comparable.
We propose a Meta Threshold Gradient Descent Regularization (MTGDR) approach for regularized meta analysis. The proposed approach is capable of selecting the same sets of influential genes across different studies, while allowing for different estimates for different platforms or experiments. To demonstrate the proposed approach, we use microarray data with binary outcome as an example in the context of logistic regression models. We analyze datasets from pancreatic and liver cancer studies using the proposed approach.
Special Notes (3/22/07):
We would like to thank you for your interest in our study. You are encouraged to try out the software. However, please note:
The software is a "research software". Shuangge Ma and Jian Huang assume no responsibility for any results produced by the software.
You are welcome to download the software and build improved version. Please acknowledge our paper or the software properly if you use the software in your study.
Please kindly inform us, if you see any error of the software.
Regularized Estimation and Biomarker Selection in Microarray Meta-Analysis Shuangge Ma and Jian Huang.
Source Code and Illustrative Examples (Microarray Meta Analysis with Binary Outcome):
Study 1 Gene expressions; Binary outcomes;
Study 2 Gene expressions; Binary outcomes;
Study 3 Gene expressions; Binary outcomes;
Source Code:V-fold cross validation.
Estimation with cross validated k.
We thank the Department of Statistics and Actuarial Science, University of Iowa for hosting the website.
For questions, comments or suggestions, email firstname.lastname@example.org or email@example.com.