Co expression scRNA-seq

Question

[tigeradab]

Is it meaningful to calculate average expression of genes from cells in a cluster (consider them as pseudo-bulk - by doing this, I will have log10 (average expression) per cluster) and using these log-normalized cluster-averages, perform a co-expression analysis (pairwise Pearson or bicor correlations) to find coexpressed gene networks that exist across cell clusters? What are the limitations to this approach?

[Wocher33]

The answer to your queston would also interest me. Did you find an answer?