About scdbfinder

Question

[carolofharvest]

Should I run scdbfinder for each sample in my data separately or should I use merged data?

Data I used is generated from 10x genomics and there are separate mouse datas from only one condition.

Answer 1

[bk11]

It will be better if you run scDblFinder in each samples separately then in the merged data.

Answer 2

[ATpoint]

You would run in a way that mimics how it was run in the lab. Doublets can only happen if the cells were processed in the same 10x droplet generation process. If you have multiple processes then run it on each separately, else run on merge.

[carolofharvest]

Thank you for your response. Is there a way to understand how the samples were processed in the 10x Genomics? Each mouse sample data was given to me as a separate H5 file. Can I assume they were processed separately ? Each mouse sample contains around 2,000 cells, and the merged data contains 14,000 cells.