Look into adding RSEM

[ewels]

From @ewels on June 19, 2017 12:5

A few people have asked about using RSEM to generate gene counts instead of (in addition to?) StringTie. It would be interesting to look into adding this either as an addition or as an alternative to StringTie.

From e-mail thread:

I was wondering whether you plan to also include RSEM for transcript-level quantification in near future (Irizarry's group has shown that RSEM slightly outperforms other methods (Teng et al., Genome Biology, 2016)). StringTie was not included in the comparison though.

Copied from original issue: SciLifeLab#132

[ewels]

From @Hammarn on June 22, 2017 11:54

Looks like there is a RSEM module available on Milou.

rsem/1.2.29

[winni2k]

Also, perhaps transcript isoform counting using kallisto/salmon might also be of interest?

[ewels]

Yes, we've loosely discussed this in the past before too. I think maybe it's best to have a separate pipeline for that though, as it's perhaps quite a different pipeline overall. What do you think? Would it work to switch out only the alignment step?

[drpatelh]

Added in #389