As expected, physical interaction datasets cluster together and are quite different from the main body of microarray expression data.
These annotation datasets are available under the Clustered data datasets, Clustered data summary and Annotations menus, as well as via searches in the SGP database.
The six clusters in Figure 4A and Figure 6 actually involve different datasets, clustered separately, that just happened to both have six clusters.
We used the datasets clustered by MCL to examine the distribution of homologues amongst protein complexes showing that they tend to be randomly distributed in both species.
Multidimensional scaling (MDS) demonstrated that the samples common to the Affymetrix and Illumina datasets cluster together and that intra- and inter-platform batch effects have been minimised.
For the taxonomically revised dataset, clustering led to 3% underestimation and GMYC to 5% overestimation (data not shown).
Figure 10 PBDE dataset clustered.
The proteins within the proteomic dataset clustered to adhesion/migration, differentiation, transport, and insulin signaling pathways.
Naturally, in this cluster of divergent contrasts, contrasts from the same dataset clustered closely together.
The microwave dataset clustered most closely to experiments in which plants were exposed to different light conditions.
This allowed us to look at how the datasets cluster – whether by platform, laboratory, experiment or otherwise.
