Performs a detrended correspondence analysis. This ordination method can be used for example in quality control and time series analysis.
The basic distinction from PCA is that PCA assumes the responses (gene expression) to be linearly related to, e.g., time, but DCA assumes the response to be unimodal (having just one peak).
Before the analysis, negative values are removed from the data by adding the absolute of the minimum values in whole data to all measurements
PNG image where the sample names are derived from the description column of the phenodata (these are editable by the user, by default they are the original file names). The samples are colored according to the group column in the phenodata