1

I'm using timeOmics to model a time series of expression data on mRNA and proteins data. The amount of samples is rather small (7 groups, 2-7 samples per group) but the features space is rather large (8k in mRNA, 3k in proteins).

TimeOmics uses the lmms::lmmSpline algorithm for spline modeling and gives a option to choose between "cubic", "p-spline" or "cubic p-spline". I could not find any hard evidence on which splince would be best for my data and when I compared them, I just find that cubic p-splines receives more features as noise compare with the other models.

Would someone advise me on which spline would be best or recommend a paper where I could find that information?

Also, I am not sure if I should set the knots as the number of samples or leave it for the algorithm to decide.

kjetil b halvorsen
  • 77,844
Sebastian Hesse
  • 111
  • 3
  • 1
    This page provides a brief summary of different types of smooths and splines. – EdM Oct 30 '23 at 02:37
  • How many distinct time points do you have? It's not clear from the question whether the "2-7 samples per group" represent 2-7 time points per group or 2-7 samples per group each evaluated at several time points. Please edit the question to clarify that, as comments are easy to overlook and can be deleted. – EdM Oct 30 '23 at 14:27

0 Answers0