What to deal with biological replicates?

Posted By: yzhang, on Dec 13, 2018 at 3:51 PM

Hi 10x friends,

In my recent single-cell gene expression experiment, I have one control and two disease samples with close (+/- 10-15% ) cells and some variation of mean reads/cell in these samples.


For the two disease samples, should I merge them then compared to the control? Or, should I compare them one by one to the control - but then I may have two slightly different conlusion?




1 Reply

Re: What to deal with biological replicates?

Posted By: 10xDave, on Dec 14, 2018 at 4:42 PM

Dear YZ



I would suggest using our cellranger aggr pipeline to combine all three samples.  By default the aggr pipeline will subsample reads from higher-depth GEM wells until they all have an equal number of confidently mapped reads per cell.




You can find more detailed documentation here: