theislab / multimil

Multimodal weakly supervised learning to identify disease-specific changes in single-cell atlases
https://multimil.rtfd.io/
BSD 3-Clause "New" or "Revised" License
18 stars 3 forks source link

Add more loss functions & ability to change the decoder output #24

Closed alitinet closed 3 years ago

alitinet commented 3 years ago

Add new loss functions for each data type:

  1. For the RNA-seq modality (see https://github.com/theislab/scarchesp/blob/master/scarchest/models/trvae/losses.py):
    • [x] negative binomial (NB)
    • [x] zero-inflated negative binomial. (ZINB)
  2. for ATAC-seq:
    • [x] binary cross entropy (BCE)

Also, for NB, ZINB and BCE, we need unnormalized counts, hence: