voutcn / megahit

Ultra-fast and memory-efficient (meta-)genome assembler
http://www.ncbi.nlm.nih.gov/pubmed/25609793
GNU General Public License v3.0
596 stars 135 forks source link

clarification on --continue #376

Open willboulton opened 2 months ago

willboulton commented 2 months ago

Hi, thanks for this great assembler.

I'm using megahit version 1.2.9 to run a relatively large coassembly - using the meta-large presets. Due to some limitations on my unversity cluster, I've set the time too small and would like to restart my job from where it left off using --continue. However, when I run this, megahit doesn't seem to continue from where it left off. I was wondering if you could clarify which parameters I should use to continue, and if I also need to change the options.json as well?

I'm currently using this:

megahit \
    --memory 400e9 \
    --mem-flag 1 \
    --min-contig-len 500 \
    --presets meta-large \
    --tmp-dir /path/to/my/assembly/megahit/samples_tmp/ \
    -t 48 \
    --12 {input.interleaved} \
    -r {input.singletons} \
    --continue -o {output}

But each time I restart, I get a new temp directory in the tmp-dir with some randomly generated hash as the directory suffix. I've also tried editing the options.json file so that continue_mode:true was set; that didn't work either. From the logs it seems like megahit is starting from k=27, even though that finished successfully... I was hoping to continue from k=47 which was where the first run left off.

Would you be able to clarify with an example how to run megahit using --continue, for example if the first run didn't use --continue and the failed halfway through?

Many thanks for your help, Will