Improve handling of rank deficient models

fgcz / prolfqua

Differential Expression Analysis tool box R lang package for omics data

https://pubs.acs.org/doi/pdf/10.1021/acs.jproteome.2c00441

MIT License

40 stars 7 forks source link

Improve handling of rank deficient models #12

Closed wolski closed 3 years ago

wolski commented 4 years ago

x$linfct_A
                               (Intercept) celltype_lncap celltype_pc3 celltype_pnt1
celltype_pc3 - celltype_pnt1             0              0            1            -1
celltype_lncap - celltype_pnt1           0              1            0            -1

 lme4::fixef(resXX$res_modelling$modellingResult$modelProtein$linear_model[[408]])
 (Intercept) celltype_pc3 
   -4.262731     3.987698

debug408.txt

currently the function my_contest returns NA for rank defficient models. Should it not be possible to somehow estimate a contrast for the conditions where measurements were observed?

wolski commented 3 years ago

Improved contrast function was implemented