It's not right to expect the number of clusters to stay the same as you're subsampling the data. Obviously, the more you subsample, the harder it is to resolve small cell populations. So the number of clusters is naturally expected to change.
Hi Jack & Christian,
I have a few questions and it would be nice to get back. Did you have the experience to work with the samples with lower cell numbers like the cells from the brain and eye? Did you any preference of the machine learning analysis for these kinds of samples?
Hi Tanima, I do have experience working with cells isolated from the brain. If you would like to discuss further, please feel free to email me, email@example.com