nanoporetech/remora

Can adapter trimming affect move table and model training?

andrewgalbraith21 opened this issue · 2 comments

Hello,

I was wondering sense barcode trimming results in discordant move tables can adapter trimming also have a similar effect especially for short (~200bp) oligo reads.

Please let me know if you'd advise on keeping adaptor trimming in this context or if it would be better to keep both barcodes and adaptor sequences.

All the best,
Andrew Galbraith

Barcoding trimming in Dorado should not result in discordant move tables. If you are observing this then I would highly recommend that you open an issue in the Dorado repository detailing this issue.

In terms of Remora, I would generally recommend adapter and barcode trimming. These features are supported in Remora given that the move table and appropriate tags are adjusted appropriate. If you are using a third party tool to perform read trimming that is not aware of the move table or other related tags then I would recommend skipping these steps since they will indeed make the reads incompatible with Remora.

Hello @marcus1487, I did barcoding trimming in dorado and resulted in around 90% of reads being discordant. I can follow up on Dorado's page about this.

Good to know that trimming is generally recommended. Thanks!