Hi everyone!
I'm trying to analyse some metatranscriptomics data from different samples, the final goal is to determine wether one group of samples has a higher expression of antimicrobial resistance genes. I'm completely new to RNAseq and I have no clue where to start. I've checked various tutorials online but they only analyse the species abundance or metabolic pathways. Could someone please point me in the right direction? A tutorial or just the names of the right tools? I've worked with this tool before https://github.com/xinehc/args_oap but applied to metagenomes, would it make sense to use it in this case? I think there shouldn't be much of a difference, just that in one case I'm analysing genes present in the genome and in the other case expressed ones. Would it make sense to use dada2 (assigntaxonomy) but switch the taxonomy database for an ARG databse, for example Resfinder?
Any response would be very much appreciated!
I don't think there is a tool that directly associates a taxon id to your transcripts using an ARG database. In your case the best strategy is to run two separate analyses: i) a low common ancestor (LCA) analysis for the taxonomic classification of your transcripts and ii) a functional annotation of your transcripts against an ARG database.
For the LCA you could try mmseqs2 or megan-lr. Regarding the second point, I don't have any suggestions.
Hi Andres,
Thanks for your answer and suggestion. I don't want to associate taxon ID to ARG, that's really complex and as you say, there are some gaps in technology (specially for metagenomics) that don't allow this. I'm specifically interested in retrieving the amount of reads that map a certain ARG database and separately those that map to 16S sequences so as to make a normalisation of ARG/16s in each sample. I believe I know an acceptable approach using WGS but I don't know if things apply also to RNAseq.
Thank you for clarifying this. I believe there is one fundamental problem with this approach.
I guess, at some point in the library preparation, you used an rRNA-depletion kit to remove the rRNA and enrich the "coding" fraction of the total RNA. You understand now that the number of 16S rRNA transcripts you find in each sample actually depends on the performances of the rRNA-depletion step rather than on the composition of the microbial community.
Regarding the normalization of metatranscriptomic data without a metagenome, you should check this paper.
This is great information, thank you very much!!