Hello,

I am using the new fitTimeSeries to test for differential abundance of genes between 4 longitudinal sampling points.  When I run fitTimeSeries, the model output for $timeIntervals says "No intervals found".  Here is my code:

Tet = fitTimeSeries(obj=MRexperiment, lvl="Class", feature="feature", class="Group", id = "Sample", time = "Time")

The "Time" variable in pData is formatted as numeric (1-4).  Also, the "Group" variable has 8 levels -- does fitTimeSeries support this many, or only 2 (the vignette only mentions differential abundance between 2 groups)?

Thank in advance for any help.

Hi Noelle,

If there are no potential regions / time periods of differential abundance estimated for a particular feature then timeIntervals slot will report "No intervals found". I am considering changing this to a more informative message if you have a suggestion.

Unfortunately, at this time only differential abundance between two factors is currently enabled, though this will potentially be extended in the future. Currently, you can either classify the samples into two meaningful groups or you can do a pairwise analysis of subsetted data.

Thanks again Joseph.  Perhaps "No statistically significant time intervals detected", or something similar.  That differentiates between the function itself not being able to identify time intervals in the pData, and the modeling process not detecting any significant time intervals.