|
HEM is a Bayesian heterogeneous error
modeling method for analysis of microarray data, developed by HyungJun Cho, PhD,
and Jae K. Lee, PhD, at the University of Virginia School of Medicine. Some
features are
- Simultaneous estimation of parameters by Markov Chain
Monte Carlo (MCMC)
- Decomposition of error variability into two error
components with experimental (or technical) and biological
replicates
- Heterogeneous error variances for all genes under
multiple conditions
- Parametric
or nonparametric Empirical Bayes (EB) prior specification of
variances
- Resampling-based False Discovery Rate (FDR) evaluation
for thresholding
- High
statistical power with a small number of replicates
Documentation:
- Cho, H.
and Lee, J.K. (2004). Vignette for HEM package in R, (PDF)
- Cho, H.
and Lee, J.K. (2004). Bayesian Hierarchical Error Model for Analysis of Gene
Expression Data, Bioinformatics, 20: 2016-2025(PDF)
- Cho, H
and Lee, J.K. (2005). Heterogeneous Error Model for Analysis of Microarray Data
Using Nonparametric Empirical Bayes Approach, submitted for
publication
- Cho, H
and Lee, J.K. (2005). Empirical Bayes Analysis of Microarray Data Without
Replication, submitted for publication
Software:
The
libraries may be freely distributed but not sold for profit. Please contact
HyungJun Cho, PhD (e-mail: hcho@virginia.edu)
at Division of Biostatistics and Epidemiology, Department of Public Health
Sciences, University of Virginia School of Medicine.
|