Descriptions of eight real-life microarray datasets are shown in Table 3.
In that respect, it does not help that the authors use only one real-life microarray study to showcase the lack of dramatic effect of technical noise.
The data that is generated using this model is more ideal than real-life microarray data, since it does not include any covariance terms.
Subsequently, the utility of this analytical toolkit is demonstrated by analyzing two sets of real-life microarray datasets from cell-cycle experiments.
This is to be done through a simulation study which involves vigorous analyses of the characteristics of predictor sets found using different values of the DDP from toy datasets designed to mimic real-life microarray datasets.
For each dataset from both collections of toy datasets and real-life microarray datasets, we have shown that there exists a value of the DDP which optimizes several characteristics representing the goodness of the predictor set.
Reapplying similar analyses to real-life multiclass microarray datasets provides further confirmation of our findings and of the significance of the DDP for practical applications.
Later in this paper, we also show how close conditions in real-life multiclass microarray datasets resemble those of our toy datasets.
After reapplying the same set of analyses to eight real-life multiclass microarray datasets, we demonstrate how the DDP works for datasets with different number of classes.
The algorithm has been applied to the promoter regions of genes occurring in 21 functionally related, co-expressed clusters, generated from Pf life-cycle microarray data by semi-supervised clustering; 34 putative cis-regulatory elements have been identified.
WSB participated in the in-life study and microarray data processing.
