Apologies for the cross-post. I accidentally put this in the bioinformatics forum first, but meant to put it here:
My group is planning an RNA-Seq transcriptional profiling study on cells taken from a humanised mouse model and we are only interested in the expression response of the human cells. We can probably sort out the human cells to 80-90% purity, but inevitably we will have some residual mouse cells and hence mouse RNA in the final samples.
I can conceive of an analysis pipeline which maps reads to both human and mouse genomes and only retains those that map specifically to human, but I'm wary of hidden biases this may impart when we generate final read counts. Presumably there may be some highly conserved genes where there is little to no difference between human and mouse and so no human specific reads.
My question is really whether anyone is aware of studies where people have done this before, or if anyone has an opinion on the best way to do the analysis. Any comments welcome really.
Alex Gutteridge
My group is planning an RNA-Seq transcriptional profiling study on cells taken from a humanised mouse model and we are only interested in the expression response of the human cells. We can probably sort out the human cells to 80-90% purity, but inevitably we will have some residual mouse cells and hence mouse RNA in the final samples.
I can conceive of an analysis pipeline which maps reads to both human and mouse genomes and only retains those that map specifically to human, but I'm wary of hidden biases this may impart when we generate final read counts. Presumably there may be some highly conserved genes where there is little to no difference between human and mouse and so no human specific reads.
My question is really whether anyone is aware of studies where people have done this before, or if anyone has an opinion on the best way to do the analysis. Any comments welcome really.
Alex Gutteridge