I am interested in the effect of an intestinal parasite infection on the rat gut bacterial community.
I have two groups of rats ("infected" and "non-infected") each consists of 6 rats. The gut bacterial community is characterised with 700 different bacterial taxa from fecal samples. Fecal samples are collected at multiple days post infection (day 10, day 20, day 40, day 70).
To answer my question, I have conducted a PERMANOVA on each days individually using Bray curtis dissimilarity index.
- Should it be better to lead my PERMANOVA with all days combined and add the variable "day" in my model ? If yes which model ? I feel a bit confused between all type of model (mixte model, ...).
- I heard about Time-Serie PERMANOVA or Sequential PERMANOVA, are these model more fitted to my question and data ?
- The variable "day" and "parasite infection" seems very correlated as the intestinal parasite life stage evolve over time. For instance, at day 10, the parasite is at the juvenile stage while at day 20 the parasite starts its adult stage. Therefore the effect of infection on the gut bacterial community may evolve over time. Does it impact the way we consider the variable "day"in my model ?
I would please like some opinions about the right way to analyse my data but also which tests are the best to use.
Thanks very much in advance Vince