A request from JohnQ:
John would like to see MeV able to display color coding for samples according to their annotation. For example, Aedin's R scripts show HCL trees with colored blocks below indicating characteristics of each sample.
I think two things would have to happen to make this work in MeV.
First: we would need to be able to sort/select samples (and maybe genes) according to annotation contents. These samples could then be put into standard MeV clusters in the cluster manager and color-coded by the user there. Additionally, if users could select groups of experiments by annotation matches, they could more easily group experiments for TTest and ANOVA analysis. This would be very useful.
Second: we would need to allow MeV's various viewers (maybe just ExperimentViewer and its subclasses) to display cluster membership for more than one cluster. Right now if a gene or sample belongs to more than one cluster only one of those memberships is displayed in the viewers. We would need to work out a way to allow the viewers to create multiple rows of color block markers for cluster membership, rather than just one.
Another thing that John suggested was adding a menu option in the ExperimentViewers that would just turn off the heatmap and show only the cluster membership and possibly HCL tree results for samples. This would make it easier to view clustering results for large numbers of samples.