|
|
Blood, Vol. 103, Issue 5, 1823-1828, March 1, 2004
Hoxa9 influences the phenotype but not the incidence of Mll-AF9 fusion gene leukemia
Blood Kumar et al.
103: 1823
Supplemental materials for: Kumar et al, Vol 103, Issue 5, 1823-1828
Files in this Data Supplement:
- Table S1. Analysis of microarray data (XLS file, 172 KB)
-
The data obtained from scanned images of the microarrays were uploaded onto an Oracle-based data server using GeneData Expressionist Suite (GeneData, South San Francisco, CA). Individual chips were assessed for quality of hybridization and other errors. The expression values were then transformed to logarithmic scale (base 2), and a linear scaling factor was applied to data from each chip. The normalized data from the chips were then compared in an ANOVA test (P < .005) to select genes that were significantly different between wild-type, preleukemic, and leukemic mice. When the genes selected by the ANOVA were used in a hierarchical clustering analysis, the samples were correctly grouped into their respective classes. A list of the genes selected by the ANOVA, along with their expression values, is included. The raw data and the ANOVA-selected list were mined for Hox genes since we were particularly interested in the changes in expression levels of this group of genes. The list selected by the ANOVA included 5 of the Hox genes: Hoxa5, Hoxa6, Hoxa7, Hoxa9, and Hoxa10. The expression values of these Hox genes in each group of mice are represented as means and standard errors in Fig. 5B. The expression values of all the Hox genes represented on the Affymetrix U74Av2 array are represented in Fig. 5C as a “hotmap” created using Treeview (http://rana.lbl.gov/EisenSoftware.htm).
|
|
