salmon on ribosome profiling data

[Zhuxitong]

Hello, I am wondering if salmon can be used to quantift ribosome profiling data(Ribo-seq) ? Because I saw that in the index step, the k-mer is 31 by default. However ribo-seq reads is usually less than 30bp.

So can I use salmon on my Rino-seq data?

[rob-p]

Hi @Zhuxitong,

You can easily change the k-mer length used for indexing by passing the desired value to the -k option of the index command. So, that part isn't technically a problem. The bigger issue is that Ribo-seq data doesn't follow the same basic model as RNA-seq data. That is, the coverage variation in RNA-seq is more often an issue to be corrected (e.g. evidence of bias during library prep / sequencing), whereas it is integral to the interpretation of Ribo-seq data (i.e. the peaks are primary features of interest). Therefore, it's not clear to me that using any RNA-seq abundance estimation software on Ribo-seq data "off-the-shelf" is conceptually the right thing to do, though you are welcome to experiment with it. However, there is some interesting work on combining transcript abundance profiles with Ribo-seq data to infer isoform-level information in the Ribo-seq data. For example, this recent pre-print provides a pipeline for this.