Rnaseq: Consider optional MMSplice for RNA splicing detection

[apeltzer]

https://github.com/gagneurlab/MMSplice

[drpatelh]

I think this should go in a different pipeline. The rnaseq pipeline is already quite bloated for standard RNA-seq QC, mapping, quantification and (differential analysis)...

Anything alternative splicing related is more suited to a specialist pipeline because its not often performed, and there are multiple options that can be incorporated there too.

[ewels]

Agreed - same thing as rnafusion and single cell analysis I think. Could be a follow-on pipeline which takes the outputs from rnaseq though?

[apeltzer]

I think we can do that yes :)

[drpatelh]

I believe you can feed most of the splice-variant detection tools the genome aligned BAM file that you would get from STAR/featureCounts or RSEM/STAR (not sure about HiSat2 because I havent used it) . So there could be a parameter in said pipeline that takes --bam as input for example which could skip the alignment steps. I suspect multiple tools will need to be implemented because most of them dont detect every type of splice-variant, and so an overview like rnafusion generates would be amazing!

For future reference, some other popular tools that could be incorporated are:
https://github.com/timbitz/Whippet.jl

http://rnaseq-mats.sourceforge.net/

https://bioconductor.org/packages/release/bioc/html/DEXSeq.html

and possibly others...

Ok, Ill close this for now then, and we can reopen and copy to said pipeline when someone decides to bring it into existence :+1: