empirical p-values for glmnet multivariate results - Githubissues

saezlab / footprints

Analysis code for "Perturbation-response genes reveal signaling footprints in cancer gene expression"

Other

20 stars 8 forks source link

empirical p-values for glmnet multivariate results #17

Open mschubert opened 9 years ago

mschubert commented 9 years ago

prerequesites

[x] map all scores to specific set of pathways
[x] use same pathway set input for glmnet

kinds of models

[x] pathways
[x] mutations and pathways as one input (to compare mut to pathways)
[x] mutations + pathways (as in, formula syntax - to compare which pathways add most on top)

generation of null models

[x] null models: 100 repetitions, shuffled labels
[x] null models: 1000 repetitions, shuffled labels
[x] real model: one repetition

calc

[x] empirical p-values for models
[ ] bar plots for best models (as defined with p-val)

mschubert commented 8 years ago

shuffling labels does not introduce enough variability - if e.g. 1 cell line out of 15 is sensitive, will fit in 1/15th of models

will not be a part of the paper, maybe look at this separately