search

mikolmogorov / Ragout

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

Lot of errors lead to "RuntimeError: maximum recursion depth exceeded" #29

Closed agroppi closed 6 years ago

agroppi commented 6 years ago

I have launched ragout with the following recipe : 

.references = persica
.target = marouch

persica.fasta = references/persica.fasta
marouch.fasta = marouch_contigs.fasta

size of the genomes around 220 Mb

The first step with Sibelia runs perfectly. But then I have a lot of errors (see full log here attached)

Any idea of what's going wrong ? ragout_Marouch_vs_persica_92577.zip

mikolmogorov commented 6 years ago

Hello,

This definitely looks like a bug. It it possible for you to send me the the ragout output directory (includes synteny blocks at "sibelia-workdir", but not the genome sequences)? This will allow to reproduce and fix the bug the fastest way.

agroppi commented 6 years ago

Hello, Here is the ragout ouput dir including the slurm script I have used to launch ragout. Thanks for your help

Marouch_ragout.zip

mikolmogorov commented 6 years ago

Hi,

Please try the latest version from the master branch - it should be fixed now. Looks like there was an unexpectedly large number of rearrangements between the reference and target genomes.

Additionally, it seems like the genomes have 70% synteny block coverage on scale 500, but only 30% on scale 5000. Most likely, this is due to contig fragmentation. If your final assembly will be fragmented (or have many unplaced sequence), consider reducing synteny block size. You can do this by adding an extra parameter to the recipe file, for example:

.blocks = 2000,500,100
agroppi commented 6 years ago

Hi, Thank your very much for your help. I have re launched the updated version Ragout according to your advices. I'll keep you informed.

agroppi commented 6 years ago

Ragout worked perfectly :) Thanks again