Bipartite Graph-based Hierarchical Clustering
Author: Calvin Chi
Bipartite graph-based hierarchical clustering performs hierarchical clustering of groups of samples based on association patterns between two sets of variables. It is developed for pharmacogenomic datasets and datasets sharing the same data structure. In the context of pharmacogenomic datasets, the samples are cell lines, and the two sets of variables are typically expression levels and drug sensitivity values. For this method, sparse canonical correlation analysis from Lee, W., Lee, D., Lee, Y. and Pawitan, Y. (2011) doi:10.2202/1544-6115.1638 is first applied to extract association patterns for each group of samples. Then, a nuclear norm-based dissimilarity measure is used to construct a dissimilarity matrix between groups based on the extracted associations. Finally, hierarchical clustering is applied.
For standard use, we recommend installing the package from CRAN via
install.packages("hierBipartite")
Minimially sufficient example of using hierBipartite. Please refer to
the
vignette
for detailed usage.
library(hierBipartite)
data(ctrp2)
groups = ctrp2$groups
X = ctrp2$X
Y = ctrp2$Y
result = hierBipartite(X, Y, groups)