Annotating unique isoforms about cdna_cupcake

[Magdoll]

Hi @kushalsuryamohan ,

(1) Please specify what Iso-Seq on which SMRT Link version (or command line) you are running? Our latest release is the SMRT Link 6.0 release and we recommend the Iso-Seq 3 analysis.

(2) Do you mean you have 2 SMRT cell and each is a different tissue (so you have 2 tissues), or you have 2 SMRT cells each coming from the same tissue?

In general, we always recommend pooling data from the same species because it increases coverage for shared transcripts and would result in higher yield. However, if you choose to continue with the current output, then combining the HQ isoforms fastq and running collapse is likely the easiest option.

For future reference, Iso-Seq 3 with Mapping in SMRT Link 6.0 already contains the mapping and collapse step provided by Cupcake. So choosing Iso-Seq 3 with Mapping in the future would eliminate the need to run Cupcake and get you directly to GFF files for annotation.

(3) The Conesa lab has developed a SQANTI software package that can compare Iso-Seq mapped output against a known annotation. The reference annotation would have to be provided in a GTF format:
https://bitbucket.org/ConesaLab/sqanti/

--Liz