Trouble Using GEO Files with Seurat

Question

[Jamie]

Hello! I am currently working on a new project, and this is my first time using seurat. I'm using data from NCBI GEO (GSE241132), and I want to perform integrated analysis. However, I am confused as to how I can create an object in seurat given there are three patients. Do I somehow merge the data together? I tried preprocessing all of the datafiles separately first, but now I'm not sure how to merge all of this preprocessed data into one object that I can then integrate. Thank you so much for your help!

Answer 1

[antonioggsousa]

Hi,

Yes, if you have three patient samples that you aim to integrate, you need to merge them first. If the three patient samples are Seurat objects you can merge them like this in version 5 of Seurat (see link):

# Merge two Seurat objects
merged_obj <- merge(x = ifnb_list$CTRL, y = ifnb_list$STIM)
merged_obj[["RNA"]] <- JoinLayers(merged_obj)

# Example to merge more than two Seurat objects
merge(x = pbmc1, y = list(pbmc2, pbmc3))

I hope this helps.

Best,

António

[1769mkc]

merge and the scanpy anndata.concat are they doing exactly same?

[antonioggsousa]

I do not know if they are exactly the same, but I think they are equivalent from its description (link). Though I should say that I am mostly an R user and I am much less familiar with anndata and Python in general. In this notebook from the Teichlab it seems to me that different anndata objects are being concatenated using the anndata.concat method. If you want to be sure, please test it and look into the object(s) before and after applying the method. This is always the best way to be sure, seeing by your own eyes.

[1769mkc]

I was trying to create a single cell object from multiple donor for single cell ATAC data where RNA seq data object is already being generated in cellxgene their portal the other part data is atac one which I need to combine from multiple donor to make for downstream integration.

[antonioggsousa]

Perhaps the best way to address your issue is to create a new post in this forum with your specific questions, along with the code you’ve used. While your question is related to this topic, you’re using anndata and Python, whereas this post focuses on Seurat and R. As a result, users familiar with anndata and Python might overlook this thread.

[Jamie]

Thank you so much antonioggsousa! You've helped me a lot!