Number of gigabases necesary for Iso-seq-PacBio

Question

[Jesús M]

Dear Biostars community, I'm currently working on a sequencing project and we are interested in using long-reads sequencing, specifically Iso-seq with PacBio to make a general survey of mRNA isoforms of Arabidopsis roots.

Considering that Arabidopsis genome is around 3.2Gbases, I'd like to know how many Gbases of data would be enough to "catch" low to moderate expression transcript isoforms.

Our aim really is to detect splice isoforms. I'd be great if you could help me with it and how to properly assess that number. I ask for the number of Gbases because the sequencing service is offering a per Gbases per sample service.

Thanks in advance!