Chuan Xu

@pcm32, layers are currently not supported; will add this option in the new version probably. Thanks for this suggestion.

@Sirin24, 1. CellTypist performs an over-clustering step. If you already have a neighborhood graph in place (i.e., from sc.pp.neighbors), CellTypist will use it; otherwise, a standard Scanpy protocol will be...

@Xyihang, there is no probability output for each cluster; but you can check the average probability and frequency of predicted cell types within a given cluster (e.g., use the `celltypist.dotplot`...

Will close the issue; please reopen it if you have further questions.

@milescsmith, can you provide more details on why this change is needed? Thank you!

@milescsmith, I think `dot_color_df` is not sliced, so their values should be matched. Do you any code to produce the scenario you mentioned?

@anke-king, if you train them separately, you will get two independent models. If you want to combine them for training, you have to unify their annotations to make cell type...

@anke-king, if the cell types in the first and second training datasets are totally different, you can combine them and train a single model. For the confidence scores, they are...

@Nusob888 1. Yes 2. Thank you for this nice suggestion although currently it is not possible (at least not directly feasible) with CellHint. I will make such function late this...

@Nusob888, please try [cellhint.selfmatch](https://cellhint.readthedocs.io/en/latest/cellhint.selfmatch.html) to harmonize different annotations for the same set of cells. This function has been added in version 1.0.0. One thing to note is that harmonization was...