Comparability of cell-cell interaction results when separately analyzing single-cell data by time periods after identical processing

[yanpinlu]

Thank you for providing such a valuable tool. I have a few questions regarding my analysis:

1. I collected datasets from different stages of the same disease and normalized them to the same scale. After running cpdb_statistical_analysis_method.cal for each of the 5 disease stages, are the interaction results comparable? How can I correctly assess the trends of cell-cell interactions across these stages?

2. If there's a significant difference in cell numbers between disease stages, will this affect the interaction results? Specifically, how does the distribution of cell subtypes impact the cell-cell interactions? Is downsampling necessary to mitigate these effects?

3. I used the microenvs_file_path parameter to separate cells by their disease progression stage, but the output did not seem to differentiate the interaction results by stage. Could you clarify how to properly use this parameter to analyze interactions for each stage?

I look forward to your insights. Thank you!