scRNA-seq analysis to compare healthy vs diseased

Question

[Chris]

Hello all,

I am still not sure about this that need your help. I have 2 scRNA-seq data: a healthy and a diseased person. So what workflow I can do to compare 2 data sets and see differently regulated genes? I created 2 Seurat objects from 2 data and perform integration like this tutorial:

https://satijalab.org/seurat/articles/integration_introduction.html

Is that correct? Thank you so much!

[jv]

yes, integrating the two data sets is the first step to making comparisons between the two.

[Chris]

Thank you for your answer. This is the opinion of another:

"The integrated object won't show a separation by disease if you use Seurat's Integration workflow. This is because batch correction would also remove any variation based on disease as well."

This made me not sure about my workflow.

[rpolicastro]

UMAPs are an imperfect representation of the underlying data, so downstream analysis should not be focused on separation of cells of the same cell type on a UMAP, but rather a more formal and rigorous approach like pseudobulk differential expression or any of the emerging perturb-seq-like analysis methods.

[Chris]

Thank you for your answer! Would you please share a sample analysis that I can follow?