RNA Contamination Tool for Developing Cell Samples

Question

[Rafael Soler]

Hello everyone,

I'm currently working on a project involving developing cell samples, and I'm encountering some challenges regarding RNA contamination. Most of the existing tools and assumptions regarding RNA contamination in single-cell RNA analysis seem to be based on adult tissue, where cell types exhibit significant transcriptomic differences.

In the case of developing cell samples, especially those in early stages or undergoing differentiation, the RNA landscape could be quite diverse and dynamic. I'm concerned that the existing methodologies might not be directly applicable or accurate for this scenario. Are there any specialized tools or approaches that could better account for the varying RNA profiles in cells during different developmental stages?

Thank you in advance for your help and input!

Best regards,

[biofalconch]

I think 10X did a compiltation of some tools that address the contamination through ambient RNA, hope they are useful :)

https://www.10xgenomics.com/resources/analysis-guides/introduction-to-ambient-rna-correction

[Rafael Soler]

Thanks! Yes, I know about this tools. The problem is that the assumptions for the actual RNA ambient correction tools are made for adult tissue, but in developing and differentiating cells, highly variant RNA can be expressed in different cell types.

[Rafael Soler]

Any more opinions?

[Ram]

Please do not add answers unless you're actually answering the top question. If you need to bump your post, you can do so in a reasonable manner using the Bump option under moderate.

[Rafael Soler]

Thanks,

I cannot find the moderate function.

[Ram]

Oops, my bad - that option doesn't seem to be available to all users. I'd recommend you request a bump from a moderator (just add a comment and tag any active moderator) instead of adding an answer.

Answer 1

[jared.andrews07]

I don't see any reason the assumptions most of these tools make would be any less valid in developing cells. If droplets with very few counts that are almost certainly empty all contain similar expression profiles, that's indicative of ambient RNA that can be regressed out of the "real" cells. It doesn't really have anything to do with dynamics between the cells at all.

From experience, both soupX and cellbender are easy to use, interpret, and incorporate into most workflows. I'd give them a try if you think ambient RNA is an issue in your data.

[Rafael Soler]

I was reading the paper carefully and I have come to the same conclusion. Thanks jared.andrews07 !