search

Nextomics / NextDenovo

Fast and accurate de novo assembler for long reads
GNU General Public License v3.0
350 stars 52 forks source link

Some bug about nextcorrect step and how can i continue the job after i change some parameter? #70

Closed AntetokounmJie closed 3 years ago

AntetokounmJie commented 4 years ago

Hi guys, i try the nextdenovo recently and i must say it really got strong power to assembly my ONT data. i wanna say thanks at first. :)

And i also found a little bug in the v2.2-beta.0

i set the parallel_jobs = 3 and correction_options = -p 30 in the run.cfg there should be 3 jobs using 30 threads per jobs meanwhile. after sort step, i got 20 correct task in total. but when it come to correct step, the nextdenovo just deliver all 20 task to my local machine. beacause the single job use about 50gb RAM, i have to close the jobs to "save" my machine. Then i continue the job mannully and i got the cns fasta finally.

So, how can i fix it? Or maybe i set the wrong parameter? :(

Besides that, i set the random_round = 20 in my run.cfg then i got the sequence with this parameter. if i wanna try 'random_round = 100', how can i excute the jobs? i really dont wanna do it start from the "align the cns fasta" step.

thanks a lot!

moold commented 4 years ago

Hi, you can use pa_correction to define the number of corrected tasks that will run in parallel. I will fix this bug in the next release and thank you for pointing out it. You can delete the file 01.ctg_graph.sh.done and chang random_round = 20 to 'random_round = 100' and re-run your main task, it will skip all finished subtasks.

BTW, keep in mind, I recommend you choose the assembly with default options as the finally result, because all other assemblies with random options may contain more assembly errors.

AntetokounmJie commented 4 years ago

Hi, you can use pa_correction to define the number of corrected tasks that will run in parallel. I will fix this bug in the next release and thank you for pointing out it. You can delete the file 01.ctg_graph.sh.done and chang random_round = 20 to 'random_round = 100' and re-run your main task, it will skip all finished subtasks.

BTW, keep in mind, I recommend you choose the assembly with default options as the finally result, because all other assemblies with random options may contain more assembly errors.

Thanks a lot, i will do it as you advised. And i am expecting the new release. :) have a nice day!

AntetokounmJie commented 4 years ago

Hi, you can use pa_correction to define the number of corrected tasks that will run in parallel. I will fix this bug in the next release and thank you for pointing out it. You can delete the file 01.ctg_graph.sh.done and chang random_round = 20 to 'random_round = 100' and re-run your main task, it will skip all finished subtasks.

BTW, keep in mind, I recommend you choose the assembly with default options as the finally result, because all other assemblies with random options may contain more assembly errors.

Oh i still got a question! i use the fast mode of minimap2 i just add the "--mode 1 --cn 20" in it. Is it right? And will the fast mode make more misassembly ? Thanks

moold commented 4 years ago

It is right and may make more misassembly

AntetokounmJie commented 4 years ago

It is right and may make more misassembly

Thanks a lot! I will try more to test it on my data. If it got marvelous result, i will send a feedback. :)