full-sized screen shots. )
Statistically significant differences in gene-gene association can be referred to as "Differentially Associated Pairs" or DAPs. The NIAID Bioinformatics and Computational Biosciences Branch (BCBB) has developed the DAPfinder and DAPview plug-ins to assist in identifying these differences through correlation and visualization.
The DAPfinder plug-in for BRB-ArrayTools uses a Fisher's Z test to compare gene-gene Pearson, Spearman or Kendall correlations between two classes and it provides an option to compare gene-gene mutual information, Pearson, Spearman or Kendall correlations between two classes with a permutation test. Several filtering options are available to help researchers identify the most relevant genes, including outlier removal options and coefficient of variation (CV) filtering.
The DAPview plug-in for BRB-ArrayTools helps microarray researchers visualize differences in gene-gene association between two classes of microarray data. The plug-in creates a XY scatterplot graph that displays the log2-expression values of two user-selected genes for two different experimental classes plotted with blue "O" and green "X" symbols, respectively. There are options to detect outliers using univariate interquartile range (IQR) and standard deviation (SD) methods from the DAPfinder plug-in. Outliers can be excluded from the graph or identified in red on the graph.
Benefits & Features
- DAPfinder computes correlations and p-values for all gene-gene pairs, producing simple text file output that can be used to generate gene association networks in Cytoscape
- DAPfinder compares gene-gene correlations between two samples or experimental groups, allowing users to compare gene association networks between two samples or experimental groups
- DAPview produces a multi-color, multi-symbol XY scatter plot that allows researchers to visualize one difference in gene-gene association between two samples or experimental groups