Microarray data analysis involves low-level and high-level analysis.
The low-level analysis focuses on how to get accurate and precise
gene expression data. The analysis built on gene expression data is
the high-level analysis such as differential gene expression
analysis, SFP detection, eQTL analysis and so on. This thesis
focuses on applications in both low-level and high-level analysis.
In the low-level analysis, the proposed L-GCRMA method combines the
advantage of the GCRMA model and the Langmuir model to get a more
accurate and precise gene expression data, especially at high
concentration. The simulation study and spike-in data analysis
demonstrates the advantage of proposed L-GCRMA model. In the
high-level analysis, a well developed SEM algorithm is successfully
applied to eQTL analysis and trait-gene association analysis.