theislab / scgen

Single cell perturbation prediction
https://scgen.readthedocs.io
GNU General Public License v3.0
255 stars 51 forks source link

Unequal size of perturbated size #72

Open XinyeZhao opened 1 year ago

XinyeZhao commented 1 year ago

Hi scgen team,

When I used the perturbation function, I found that the size of the predicted data is not consistent with the ground truth data, for example, the stim1 data I used had 3394 cells, while the perturbated output only has 3333 cells, and I traced back in the source code and I found the function named "balancer". Can you tell me what's the purpose of using the balance function?

Thanks