Reasoning for Aggregation by Summation

[DarioS]

aggregateAcrossCells uses summation rather than averaging. Is the reason for that the fold change preservation of summation?

Source: Supplementary Figure 5 of Muscat Detects Subpopulation-specific State Transitions from Multi-sample Multi-condition Single-cell Transcriptomics Data, Nature Communications.

If so, it might be helpful to have a \url link in the function documentation to guide novice users to the reason.

[LTLA]

The motivation was not so sophisticated. Summation is just to get raw counts, which is needed for downstream analyses with edgeR, voom, etc. With averages, the precision of each observation depends on the number of cells in each group; this information is not easily introduced into the weighting frameworks of various DE methods.