Entering edit mode
4.2 years ago
getanid123
•
0
Hi all,
I was trying to run abyss 2.2.4 and I came across an issue. The loop includes running 5 kmers on 20 samples. The code used is as follows:
abyss-pe name='name' k='kmervalue' in = 'file1_r1.fq file1_r2.fq'
My loop is running fine and execute all 100 samples. But for 5 samples in between [not the extreme kmers or extreme samples in the list given to loop], some output files are not generated. I was wondering if I need to specify the thread parameter for abyss?
Please suggest!!
-Getan
What kind of output files are missing?
apart from the fact that for instance with certain kmers (or input) it can't get to an assembly result it should do the same for all samples.
I was looking in particular for scaffolds and contigs files and they are not generated. But I was able to generate these files after running it again using same command line [outside loop].
so this issue is (somewhat) resolved thus?
It is a big task if I have to check all the samples for contigs.fa and regenerate all the missing contigs. I just wanted to know if I need to specify -np or -j parameters. I want to assign specific processes or threads to be on safe side.
To clarify, each of these runs were done on a different directory? How much memory does your system have? And do you have the log files for the runs that fail. The thread parameter shouldn't matter as it will be local to each run.
Hi john, Yes the output files were saved on a different directory. The system has 512GB and 16GB RAM. The output file [.dot, .fa, .hist, .path, .path.dot ] are generated but I kind of deleted them to go for a new run [which generated all the data I needed]. But would appreciate if you can guide me as to which of these files in particular should I be looking at to understand the fail of some runs. Thanks!
Can you go to the directories where the runs failed and rerun your command, and log the output.