I have a maker file that has 30 cluster and a dataset with 38,000 cells. However, when I tried to train the classifier with this dataset, in the error message it says that I don't have enough cell for any cluster. If I use less cluster in the marker file (~10), it will work. May I ask what can I do in this case?
And can I use the same dataset for both training dataset and classifying dataset since it is very difficult to find similar exiting dataset? Technically it's possible but what would be the potential disadvantages?