Enter the body of text here

Hello,

I have 1 "file" that contains scRNAseq data from cells derived from 3 mice, each labelled with one CMO. The three mice same condition (PBS), where labeled (CMO) and pooled and there was one seq run (1batch, am I right?).

I demultiplexed the 10x data using cell ranger.

I'm using bioconductor following this book/guideline: https://bioconductor.org/books/3.15/OSCA.multisample/human-pbmcs-10x-genomics.html#quality-control

My question is: should the tSNE from biological replicates look somehow similar? I'm thinking in shape but also number of clusters. How can I be confident that cluster changes after treatment are due to treatment if there is such "variability" in my PBS triplicates?

I hope the question makes sense

Sorry if this has been questioned before Cheers Jose G

You shouldn't expect t-SNE plots from different mice to look similar. Simply re-running t-SNE on the same sample will give different results, let alone on different samples. Note that the three samples in the book chapter that you are following don't look similar.