For Oxford Nanopore contigs produced by any long-read assembler has anyone performed any benchmarks to compare the polishing tools racon, nanopolish, or something like wtpoa-cns?

The compute time of nanopolish is quite long and I would rather use another tool if it will suffice. After long-read polishing I will polish more with two rounds of pilon using 200x coverage Illumina shotgun data from the same individual.


Nanopolish is necessary if you want to get high-quality consensus. racon etc don't use signal data. They can't achieve high quality. pilon at times doesn't work well. Even if it worked perfectly, it wouldn't help repetitive regions inaccessible by short reads.

  • $\begingroup$ You're quite right. I polished twice with racon then twice with pilon. Mapped short read pairs and long reads. All of the gaps in the short read mappings were covered with long reads. Going to use nanopolish now! $\endgroup$
    – conchoecia
    Mar 5 '19 at 18:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.