Finding players and targets of expression-regulating pathways

Question

[Tobi]

I am currently working on a project involving using pathways to group genes. However, I am quite new to the field, so I have a some (beginner) questions concerning the biology and available databases.

A) I have heard about the concept of dividing the genes interacting in pathways into players, i.e., the corresponding proteins interact, and targets, i.e., those genes whose expression gets regulated. For my project, such a distinction would be very helpful. Is this a common concept known under different nomenclature? Searching for the terms “players” and “targets” I only found individual pathways. Databases like KEGG, Reactome or PathwayCommons don’t seem to integrate this concept. Are there other databases like these using such a division of interacting genes?

B) As I did not find any such database, I thought about using gene sets from KEGG and Reactome and looking up their gene expression regulation targets using the sif file(s) provided by PathwayCommons. I thought about two ways to get the grouping:

• dividing the gene sets into “player” and “target” genes based on the sif file retrieved from PathwayCommons by looking up the “controls-expression-of” relation
• looking for further targets by traversing the graph given by the sif file of all interactions in PathwayCommons for a few steps, starting at all genes involved in the pathway. I just tried it for two steps, giving me already quite long lists of genes, so I am not sure if this is a reasonable approach. My main idea behind the second approach was that many pathways have only few genes listed as members regulated by it (in the order of tens). This contradicts my understanding that the number of targets should be in the order of hundreds.

Do you think that this is a legitimate approach to the problem? Is there already a solution to it, which I might have overlooked due to different nomenclature? If you spot a misunderstanding on my side, could you please explain it or point me to resources, so I could clarify things up?

Thank you in advance for help with any of these questions!

[Pratik]

Hello,

So I am unsure if this is exactly what you are looking for. But there are tools that can identify ligand-receptor interactions (cell-cell communication) using scRNA-seq data. InterCom is one in particular that comes to mind, which I think I saw uses curated gene lists for signaling pathway identification internally: https://github.com/saschajung/InterCom which can help you build a heatmap like this:

Image Citation: Gonçalves, C.A., Larsen, M., Jung, S. et al. A 3D system to model human pancreas development and its reference single-cell transcriptome atlas identify signaling pathways required for progenitor expansion. Nat Commun 12, 3144 (2021). https://doi.org/10.1038/s41467-021-23295-6

Another tool, that I was shown recently is rPAC: https://doi.org/10.1016/j.ymeth.2021.10.002 which I believe uses the gene lists from KEGG pathways internally to identify (score) which pathways are up-regulated in RNA-seq data, this could probably be adapted for scRNA-seq data as well?

I think there are other more popular/used tools that accomplish the same things above. I just can't remember the name of one in particular that I saw mentioned in some papers. (I think this one was the one I was trying to remember when I made this post: https://www.cellphonedb.org/ )