Multi-Condition Gene Expression Data

Next: Tissue Gene Expression Data Up: Clustering Using BioClust Previous: Temporal Gene Expression Data

Multi-Condition Gene Expression Data

The gene expression data used in this analysis is from [6]. Kim et al. studied gene regulation mechanisms in C.elegans.

Gene expression patterns were measured for 1246 genes in 146 experiments, resulting in expression matrix. The similarity matrix was obtained using Pearson correlation.

The algorithm found 40 clusters. Only very few genes out of the 1246 were classified into families by prior biological families. The algorithm clustered these families quite well into few homogeneous clusters. Figure 12.7 shows the expression patterns #1 which contains all the known sperm proteins out of all the 1246 genes.

**Figure 12.7:** Expression patterns of cluster #1. This cluster contains all the known sperm proteins out of all the 1246 genes.
$\resizebox*{5in}{4in} {\includegraphics{lec12_fig/resKim1.eps}}$

One example of the potential use of clustering for analyzing gene expression patterns is shown in figure 12.8. A six-gene cluster containing two growth-related genes and four anonymous genes. This suggests the possibility that the other four genes are also growth related.

**Figure 12.8:** Expression patterns and members of cluster #24. This cluster contains two growth-related genes *lim-15* and *E2F*. This suggests the hypothesis that the other four members of this cluster have related function.

Next: Tissue Gene Expression Data Up: Clustering Using BioClust Previous: Temporal Gene Expression Data

Peer Itsik
2001-02-01