Skip to main content
Figure 7 | BMC Bioinformatics

Figure 7

From: Quadratic regression analysis for gene discovery and pattern recognition for non-cyclic short time-course microarray experiments

Figure 7

Examples of genes with similar expression patterns in terms of mean signal and regression. a. Sfpi1 is an example from k-means cluster K2; b. Anxa2 is an example from k-means cluster K8; c. Clu is an example from k-means cluster K5; d. D17H6S56E-5 is an example from k-means cluster K6. a and b are examples of genes with the same up-down-up-up pattern (up-regulated at the second time point, down-regulated at the third time point, then up-regulated at the last two time points) in terms of mean transformed signals (green lines). They also have the same LU regression pattern, but are in different k-means clusters. c and d are examples of genes with the same down-up-down-up pattern in terms of mean transformed signals (green lines). They also have the same QLVU regression pattern, but are in different k-means clusters. Clearly, the regression method provides better classification of the temporal expression profiles of these genes than the k-means clustering method. The horizontal axis is the log transformation of time. The blue dots are the signals. The green line is the connection of the mean signal at each time point. The red line or curve is the fitted regression pattern.

Back to article page