cancer immune subtypes based RNAseq data

Question

[supertrouxa1234]

I am bioinformatics beginner and immune genomics enthusiastic. based on a recent paper http://www.cell.com/immunity/pdf/S1074-7613(18)30121-3.pdf I´d like to apply the cancer-immune subtypes C1–C6 classification to my RNAseq data, and would need a hand on it.

Answer 1

[Kevin Blighe]

You have coincidentally chosen a publication for which very detailed and comprehensive methods covering multiple different programs has been made available. The authors also speak of manual curation of their signatures by 'expert' immunologists. I do not believe that anybody here has the time to take you through this step-by-step, and I am replying now because I doubt that anybody else would have [replied].

In light of that, my recommendation for you is to go the section entitled 'METHOD DETAILS' in the publication and to start from line 1 of that section. When you hit something that you don't know, learn it, and use BioStars of course for assistance by searching for pre-existing questions. The best way that you will learn is by not trying to do everything at once.

Kevin

Answer 2

[ivivek_ngs]

Well adding up to what Kevin has said, if you are new, learn and unlearn to get comfortable. I would not make much of a details, just a bit of cue that should be good enough for your starting point. Try to frame your query as to why you want to use the subtypes and what you intend to address with that.

Questions you should ask:

1. Does your data set have enough sample power? If so then definitely you have more things to achieve. Just hit on the supplemental data , you should have the gene lists for each of the subtype. Just extract them and throw them in your data and see their behavior in your data and enrichment. Are these subtypes all statistically relevant or if any of one/few subtype(s) has more prominence?

2. Well what clustering methods should I use or exploratory visualization? One more query for you. A lot can be done by mere clustering methods and enrichment followed by functional relevance of those.

Well I gave enough for you to start. Now read, seek Biostars/SE other blogposts about how to play the magic with R/python or custom tools to seek what you intend to but first read enough literature to formulate your queries. Good luck!