I am a bit confused about a painful step of my pipeline using Sleuth. I am trying to move from the transcript level to the gene level and that's why I am using biomaRt.
PS: please pardon my ignorance in term of bioinformatic jargon
Basically, there is some weirdness between some of the ensembl annotations. Your transcriptome probably says something like 'ENSMUST00000148678.2' for that specific transcript. When we try to merge that with the target_mapping, it ends up being empty because the target_id is 'ENSMUST00000148678' (note the missing .2). It's basically just telling you that we removed .N and it seems to have fixed it. When we do that and there is no intersection mapping still, then the program will completely error.
If you still get results after running a test and calling sleuth_results(), then everything should be fine.