search

mikolmogorov / Ragout

Chromosome-level scaffolding using multiple references
Other
149 stars 27 forks source link

Input is larger than 1 GB, can't proceed #58

Closed fergsc closed 4 years ago

fergsc commented 4 years ago

Hi, I have assembled 21 plant genomes of ~600mbp and would like to use Ragout to scaffold them. I have sucessfully run sibeliaZ on my genomes and produced a maf file of 14G. When I try to run Ragout I get the error error: Input is larger than 1 GB, can't proceed. Is there a hard limit on the input data size for Ragout and if so is there a way around this issue?

thanks

mikolmogorov commented 4 years ago

@fergsc SibeliaZ has not yet been integrated into Ragout. The original Sibelia can't handle input that large. The current solution is to use progressive cactus alignment. I also would recommend to select 4-5 closest references instead of using all 21.

Mikhail

fergsc commented 4 years ago

Thanks for the quick response.

Based on this https://github.com/fenderglass/Ragout/issues/40 I thought sibeliaZ was capable of being used with ragout. I shall try and convert my maf into a hal and see how it goes. Running Cactus in my HPC is hard/not possible. Was really hping to use Ragout, all my genomes are very closly related and have a lot of shared synteny.

Thanks again.