SC EPSCoR/IDeA

South Carolina Experimental Program to Stimulate Competitive Research and Institutional Development Awards
Home   About Us   News   INBRE   EPSCoR   COBRE   Funding   Evaluation   Contact Us   Citation Information  

Agilent DNA Microarray Data Analysis

Supported by the SC INBRE Biotechnology Core
Pre-analysis:

Researchers are advised to consult personnel of the SC INBRE Biotechnology core to discuss Agilent microarray experimental design and costs.  To initiate an Agilent array analysis project, please contact either Dr. Pierre Rivailler at pierre@biol.sc.edu or Dr. Kim Creek at Creek@med.sc.edu.

Analysis:

1. Background correction using normexp algorithm.
2. Normalization within arrays using lowess method.
3. Compute overall intensities (A) and log ratios (M) values.
4. Create new parameterization based on experimental design.
5. Identification of differently expressed genes using either a) fold change (CARMAweb) or b)
   statistics (empirical Bayes moderated t-statistic, the corresponding P-values adjusted to control the
   false discovery rate and empirical Bayes B log odds of differential expression) generated by
   limmaGUI.
6. Significance analysis of gene ontology (GO) terms associated with differentially expressed genes.
7. Hierarchical clustering of differentially expressed genes and/or samples.

Output Provided to the Client:

1. An Excel file of normalized overall intensities values (A) and annotations for all genes.
2. An Excel file of normalized log ratios values (M) and annotations for all genes.
3. An Excel file of differentially expressed genes containing gene expression data, filtering metrics and
    gene annotations.
4. Three Excel files reporting GO term significance analyses (biological process, molecular function
    and cellular component).
5. An image file (jpg) of clustering analysis.

Other Analysis Upon Request:

1. Replication of above using different filtering criteria.
2. Correlation analysis to find genes with specific expression patterns.
3. Sample cluster analysis to explore how different clustering/distance algorithms define sample
    relationships.
4. Training in the use of limmaGUI.
5. Training in the use of carmaweb.