search

wwood / finishm

genome improvement and finishing without further sequencing effort
MIT License
5 stars 2 forks source link

I got an error on 'roundup' command #39

Closed minsookim1983 closed 8 years ago

minsookim1983 commented 8 years ago

Hello,

I am trying to do 'roundup' with contigs, but I got an error that I attached. error.txt Please, give me any way to solve this problem. Thank you.

Min-Soo Kim

wwood commented 8 years ago

Hi Min-Soo,

Hmm, there seems to be some problem running the velveth that comes with finishm, which I would guess is due to some issue with the input files. Would you mind running the following commands and reporting back the results please, from the same directory where finishm was run?

file d6f32d9d-7ba0-42ed-97ea-d79eb91abdae/msk2/lysogen_data2/bacteriome.merge/all.screened.mm10.fa.pair.R1.fq
file d6f32d9d-7ba0-42ed-97ea-d79eb91abdae/msk2/lysogen_data2/bacteriome.merge/all.screened.mm10.fa.pair.R2.fq

head d6f32d9d-7ba0-42ed-97ea-d79eb91abdae/msk2/lysogen_data2/bacteriome.merge/all.screened.mm10.fa.pair.R1.fq
head d6f32d9d-7ba0-42ed-97ea-d79eb91abdae/msk2/lysogen_data2/bacteriome.merge/all.screened.mm10.fa.pair.R2.fq

thanks.

minsookim1983 commented 8 years ago

Here is the results of the commands that you gave me. error.txt As the results, something is wrong with the input files.

Min-Soo

wwood commented 8 years ago

It seems those files don't exist, or I'm not interpreting things correctly. How are you running finishm? i.e. what command line?

minsookim1983 commented 8 years ago

I am sorry to bother you. I just found what was wrong with me. As you indicated, the input file path was wrong. I corrected it, and it is working now. Thank you.

wwood commented 8 years ago

Nice easy problem then, no worries.

minsookim1983 commented 8 years ago

One thing more that I want to ask is that, Is there any way to set cpu number? I did not find any flags for cpu number in Usage. I checked issue #4, and I guess it is still not adopted to finishM.

wwood commented 8 years ago

Currently only the velvet step is multithreaded and in my experience much of that does not multithread. However, it can be useful by setting the environment variables as velvet describes in its manual (as in #4). I really should fix #4 though..