Multivariate analysis of obese and lean subjects based on gene expression signals. Principal component analysis (PCA) was performed on lean and obese subjects based on 12128 Affymetrix probe-set signals. A scatterplot of the first two principal components demonstrate a general separation of the obese and lean phenotypes along the first principal component (PC1). Model parameters are as follows: Further details on the PCA model parameters are included in Supplemental Table 2.
Ghosh et al. BMC Medical Genomics 2010 3:56 doi:10.1186/1755-8794-3-56