So i have a question regarding basecalling with Dorado. Say i start with a POD5 file from a sequencing run and I now want to basecall it with a sup model I already have downloaded, demultiplex and trimm all adapters and barcodes.
I find it somewhat difficult to make out from the documentation if you would need to run 2-3 seperate Dorado commands (Dorado basecall and Dorado demux, potentially with a third one for trimming?), or if you could do it all in one command. Because if I look at the most recent documentation, especially regarding the trimming (see link below at the bottom), I would read it as "If you would run demux seperately, do not trim in basecalling". But would this then also not imply that in a single Dorado command, you could perform basecalling, demux and trimming like the command in the "Effect on demultiplexing" section:
dorado basecaller <model> reads/ --kit-name <kit-name> --trim all --emit-fastq --output-dir dir1
https://software-docs.nanoporetech.com/dorado/latest/basecaller/read_trimming/
Would the above command for example result in trimmed, demultiplexed reads in fastq format, with a file per barcode?
Any input from more experienced users is greatly appreciated!
1 answer
as I understand it, the command you provide will do basecalling & trimming (== why the kitname is provided) but not demultiplex.
For the demultiplexing you need to run the demux command separately.
What you command will do is to classify the barcodes (== add tags in the bam output file) but will NOT split the barcodes out in a file per barcode . For the split files you will need to run the demux step
but depending on what you want/need several approaches are valid.
The only thing to pay attention to (== a big DON't :) ) is to do the basecalling with trimming (--trim all) without providing the kitname (for the barcoding) !! In this case you will loose the barcodes and you can't recover them anymore (re-basecaling is then your only option)
Log in to answer this question.