Association with subject-level continuous phenotypes

[hruffieux]

Hi,

Thanks for the great package!

Does diffcyt allows for settings where estimating associations with a continuous phenotype at the subject level (e.g., BMI of subjects) is of interest, or does it only consider proper differential analysis settings (comparing groups of subjects, e.g., overweight vs normal weight) ? In the latter case, is there any reason why running an LMM for a continuous phenotype wouldn't be appropriate in a flow/mass cytometry context (I'm new to the field!)? Would this be because of statistical power concerns if the number of subjects is low? If so, I guess your functions could be easily adapted to handle this scenario (for applications where the number of subjects is reasonably large)?

Thanks !