Application to bulk-seq data?

[hsiowa2]

Hello.

First of all, thank you for developing a very nice tool. I have two questions:

1. Is it possible to apply pagoda2 to bulk-sequencing data as well?

2. It seems in pagoda2, hierarchical differential expression function was added compared to the original pagoda's overdispersion gene sets in Nature Methods paper. Could you give brief explanation about the difference, apart from the speed described in the tutorial?

Thank you in advance.

[evanbiederstedt]

Hi @hsiowa2

Apologies, I keep forgetting to respond to your questions.

Is it possible to apply pagoda2 to bulk-sequencing data as well?

Bulk RNAseq and single-cell RNAseq have very distinct use cases in genomics. It's difficult to envision why you would do this. Pagoda2 was specifically designed for single-cell analysis.

It seems in pagoda2, hierarchical differential expression function was added compared to the original pagoda's overdispersion gene sets in Nature Methods paper. Could you give brief explanation about the difference, apart from the speed described in the tutorial?

The original tool is here: https://github.com/hms-dbmi/scde

Pagoda2 included some statistical methods from the original papers, but other functions were written. I think you're referring to this method getHierarchicalDiffExpressionAspects().

https://github.com/kharchenkolab/pagoda2/blob/main/R/Pagoda2.R#L634-L785

Reading through the R code and Roxygen2 documentation should answers all of your questions, I hope. That's the best explanation possible.

Best, Evan