Dev

[derrik-gratz]

generalize PCA, add some new normalization methods for heatmaps

[derrik-gratz]

@micahpf @gktharp1 drawing your attention to this as it changes how the heatmap function works a little. It now just expects a matrix of input values (not necessarily a DESeq object) so this is more generalized for other datatypes. It also pulls the baseline subtraction component out of the heatmap function intonormalizeDataForHeatmap, which has the similar functionality the heatmap used to (provide a grouping variable and a baseline level). So you normalize the data outside of the heatmap (if you so choose), then pass it into the heatmap. There is also normalizeDataForHeatmapByIndividual which is good for timeseries data, allowing you to subtract an individual's baseline from it's other samples.

Example of normalized by individual I kept the baseline samples in to be demonstrative, but they can be omitted from the return of normalizeDataForHeatmapByIndividual.

See the previous output here, where it could only be normalized by group. This is still possible using just the normalizeDataForHeatmap function

[derrik-gratz]

You don't have to review this (you can if you'd like). I'll push to main at the end of the week