search

ncborcherding / escape

Easy single cell analysis platform for enrichment
https://www.borch.dev/uploads/screpertoire/articles/running_escape
MIT License
143 stars 20 forks source link

NES comparison between cells #23

Closed Jeron1996 closed 3 years ago

Jeron1996 commented 3 years ago

Thanks for this great package!

I have a question regarding the comparison of genesets within a cell. I want to use the C8 collection and determine which of those gene sets is the most enriched in a particular cell.

Currently, I am doing a z-transformation on the EnrichIT output for each gene set and then determine which gene set has the highest z-score within a particular cell. Could you comment on this approach and perhaps advise on a more streamlined method?

Best,

Jeron

ncborcherding commented 3 years ago

Hey Jeron,

Thanks for reaching out.

One issue I have with the approach you describe is comparisons at the level of a single cell. Due to drop-out effect, we expect to have a degree of heterogeneity among enrichment of cells derived from the same origin. When you are thinking about enrichment in single-cell, I would encourage the analysis at multiple cell levels or cluster level when possible.

So one thing I do is find the median enrichment of gene sets of interest by cluster or other categorical variable (like tissue) and just look at the data as a heatmap.

Hope that helps, feel free to reach out if you have any more questions! Nick

Jeron1996 commented 3 years ago

Hi Nick,

thanks for clearing this up for me. Now I also understand the GSEA approach in your clear cell renal carcinoma paper a lot more.

Cheers, Jeron