Question: Loupe Browser (v7.0+) enables differential expression within cell types and between experimental conditions with multi-sample comparison (aka pseudo-bulk). How many replicates are needed in order to detect differences in my comparison of interest?
Answer: The statistical power of your experiment depends on a number of factors including, but not limited to:
- Sequencing depth
- Sample size
- Total number of cells per individual sample
- Number of cells per cell type of interest
Because each experiment and its goals differ, we do not have a general recommendation on the minimum number of replicates needed to achieve your experimental goals. We recommend estimating the power of your experiment during the experimental design process to determine whether you have sufficient power to detect differences in your comparison(s) of interest given a certain sample size. You can calculate power for single cell multi-sample experiments using a tool like scPower. For more information about power analysis in multi-sample comparisons and general considerations for single cell analysis, we recommend this paper:
Regan C, Preall J. Practical Considerations for Single-Cell Genomics. Curr Protoc. 2022 Aug;2(8):e498. https://doi.org/10.1002/cpz1.498. PMID: 35926125; PMCID: PMC9479272.b