Nextomics / NextPolish

Fast and accurately polish the genome generated by long reads.
GNU General Public License v3.0
213 stars 28 forks source link

[question] combine with medaka or not? #114

Closed JWDebler closed 1 year ago

JWDebler commented 1 year ago

In the past my long read polishing pipeline was:

1 x racon followed by 1 x medaka

I was wondering if you have any data on potential combinations of polishing tools.

Is there anything nextpolish does better/worse than other polishing tools? Does it make sense to combine them or run nextpolish only? Cheers, Johannes

moold commented 1 year ago

If you don't have short reads, I think medaka (long) is ok. If you have short reads, you can try medaka (long) + nextpolish (short) or nextpolish (long + short reads)

JWDebler commented 1 year ago

I ran a quick polisher combination comparison. In this case it looks like nextpolish followed by medaka (using long reads only) is the winner by a tiny margin.

image

moold commented 1 year ago

Nice benchmark