Hi so I've seen this mentioned here and the github. However this instance seems to missed by those answers. I was running with an Ensembl genome and the default miso genes. However sashimi failed and it was due to the pickle index not recognizing the gene. I tracked it to the work dir entry and sure enough couldn't find the ENG0000000nnnn match. However, I did find the ENG0000000nnnn.1 id. I ran the same files through but used the igenome for human assuming that would definately contain the default miso genes used by the pipleine. However, I go the same error but a different gene. So third time I looked into miso a bit and found out you could select your own gene IDS. So I went back to Ensembl, looked in the gtf and selected two that were definately in there --miso_genes ENSG00000226950,ENSG00000250302. Still failed on Exception: Event ENSG00000226950 not found in pickled directory index. Now I hunted down into the work folder again and sarched the pickle index. It had an entry for ENSG00000226950.8 So it looks like pickle index appends a "version" to the ENnnnn so miso/sashimi can't see it. Any idea how I can get this to work? Not ideal but if i now go in and manually edit the --miso gene name will it acctept something with a nnn.n in it? could I edit the pickle index? will it throw out eveyting else ie not match somewhere and throw a wobbly?

You can select your genes of interest with the

--miso_genes

flag. As you said, the pipeline adds an extension to the ENSG id. That is why you have to go to the index directory of the specified gene and add it to the name that you had previously specified. I´ll give you an example. Gene of interest: ENSG00000226950 Command (solution):

--miso_genes ENSG00000226950.8

Thanks Sergio. Its a faff but definitely works.