Is the cell type label a must?

[yzcv]

Hi,

I am confused that why scgen needs to feed in the data with cell type labels. Does that means your scgen is a supervised learning method? Hoping your reply. Thank you.

[M0hammadL]

Hi, It does not need cell type labels for training, However, in the prediction step, you need to pass a certain cell type to predict function to transform it into your desired condition. please have look at this example. cell type labels are only used to make the computation of the delta vector balanced.

[yzcv]

Thank you for your reply. I am sorry I did not make myself clear. I am quite interested in your method scgen applied to integration of datasets. In this example, the VAE network needs the cell type labels as a part of input. Does that mean batch removal based on scgen is a supervised learning algorithm as it need the cell type labels? As your method extremely outperforms other batch removal methods which do no need labels, I want to make sure whether your method uses additional information except for the gene expression matrix. Hoping your reply. Thanks a lot!

[M0hammadL]

Hey, sorry for the late reply:) yes it the case of batch removal it can be considered as supervised with respect to batch labels.

[yzcv]

Hi, can this method be considered as supervised with respect to cell type labels ?

[M0hammadL]

yes:)