Kenneth Lim

Hi @warrenmcg, I've run into the same issue (Transcript level analysis runs without error while gene-level aggregation fails). ``` > so = sleuth_prep(s2c, md, target_mapping = t2g) reading in kallisto...

I've done a quick followup for my case. Here, the issue seems to lie in the `reads_per_base_transform` function in measurement_error.R. In particular, this seems to stem from the following bit:...

@warrenmcg I've taken a closer look at the mappings and have noticed the presence of duplicate (target_id, ens_gene) pairs stemming from the allocation of multiple gene names to the pair...

@rajas2211 So, the problem is that this results in odd behavior you'd see python fallback to those, breaking the point of the separate venvs to begin with. The st-conda package...