Input data from pacbio single cell long-read

[leewqa]

Hi SCOTCH team,

I was wondering if SCOTCH has been tested on Pacbio single cell long-read data? I would like to try SCOTCH for isoform quantification of a Pacbio single cell long-read data set and have a question on what is the appropriate input data for SCOTCH.

I had previously ran a preprocessing pipeline and obtained deduplicated mapped bam file. Can that be used as the input data?

Thanks a lot for your help!

[Karenxzr]

Hi, SCOTCH has been tested on pacbio single cell long read data. In the manuscript, I used isoseq for preprocessing the input (tag, refine, and dedup). Generally speaking, the input file required by SCOTCH is bam file that has been aligned to genome with cell barcode and umi tags, which can be set up using --barcode_cell and --barcode_umi.

[leewqa]

Got it. I am trying to run SCOTCH annotation and should I be using --reference_pkl data/geneStructureInformation.pkl, if I am using my own reference file i.e. --reference gencode.gtf?

My understanding is that by setting --reference_pkl None, main_preprocessing.py will generate the pkl file using the input .gtf but I might be understanding this wrongly. And when I set --reference_pkl None, I am running into a NoneType error

[Karenxzr]

You are understanding it correctly. you can set up as --reference_pkl 'None' --reference 'path/to/reference.gtf'

[leewqa]

Thanks. I am running into TypeError: argument of type 'NoneType' is not iterable error when setting --reference_pkl None. I can send the full error over email if that helps. Thanks!

[Karenxzr]

please send me the annotation log file and the full error message. thank you! cyranvvv@hotmail.com