search

satijalab / seurat

R toolkit for single cell genomics
http://www.satijalab.org/seurat
Other
2.29k stars 915 forks source link

Which assay(integrated vs RNA) to use for importing batch-corrected data to monocle3? #1674

Closed Sophia409 closed 5 years ago

Sophia409 commented 5 years ago

Hi,satijalab! Should I use the batch-corrected, integrated gene values in the "integrated" assay or the non-integrated gene values in the "RNA" assay for downstream Monocle analyses? And As you said in the past,Seurat does not return batch-corrected expression values ,Instead, Seurat can be used to generate a corrected cell/cell distance matrix. . Does it mean that integrated gene data matrix values in the "integrated" assay is a corrected cell/cell distance matrix? If so,we couldn't import the integrated data to monocle.Can we only use monocle to do batch correct and further pseudotemporal ordering analysis ?Has any way to import the batch corrected data integrated by seurat to monocle and do further analysis?

satijalab commented 5 years ago

Thanks for the question. Please note that for our previous answer, that was for Seurat v2, and does not hold in the new version.

You should use the intergrated data for constructing a pseudotime trajectory.

However, if you are doing any comparisons across conditions (for example, if you construct an integrated trajectory for WT and KO cells, and want to find genes that change with pseudotime in the WT but not in the KO), you will need to do those comparisons on the original data.

Sophia409 commented 5 years ago

@satijalab Hello, thank you for your explanation. I'm still a little confused.You mean we should use the intergrated data for constructing a pseudotime trajectory.But there are only data slot and scaledata slot on the integrated assay.The batch corrected counts slot doesn't exists, which is the key to import to monocle to do analysis.And monocle people advise to give the counts slot not the normalized data.when you give it the batch corrected integrated assay,actually you give a corrected cell/cell distance matrix. That is pretty tricky for monocle to analysis. How can i deal with the problem?

veravevo commented 5 years ago

I'm having the exact same problem: I have used integrated Seurat-data to construct the trajectory, but need to used the "RNA"-counts to perform comparison between conditions in Monocle2. However, only one assay can be stored in the CDS, and changing values in the CDS is not possible. Is there a solution to this issue?

rg734 commented 8 months ago

I was just wondering whether there was ever any resolution/additional thought around this problem?