Polishing ONT with HiFi vs Illumina #146

Hello,

Thank you for creating such great software!

I am in the planning stage of assembling the genome of an autotetraploid. I will be generating approximately 50× ONT ultra-long reads and 50× Pore-C data using a PromethION 2. I also have enough budget for either 10× PacBio HiFi data or 30× Illumina data.

Because I am working with a polyploid, I like the idea of using HiFi reads, since their read length may be helpful for resolving homologous chromosomes and duplicated regions. However, it seems that most ONT polishing/error-correction workflows either use Illumina reads or substantially higher HiFi coverage.

Do you think NextPolish would provide a notable improvement to an ONT assembly using only ~10× HiFi data? In particular, would you expect 10× HiFi to be preferable to 30× Illumina for polishing a 50× ONT assembly, or would the lower HiFi coverage make it less useful?

Thanks so much for your input!