Hello,

weijenc @ SEQanswers a écrit :

> Dear Seb567,
>
> I have been trying to use Ray to assemble SRR131208 (containing paired-end, 
> 100 bp Illumina sequencing data). I have been

>testing using 4 nodes x 8 processors, and 8 nodes x 8 processors with kmer = 
>21. The 4 x 8
 > is still running, while the 8 x 8 kept generating error messages:
> .
> .
> .
> Rank 35 reached 400 vertices from seed 73, flow 1
> Speed RAY_SLAVE_MODE_EXTENSION 2320 units/second
> Rank 35: assembler memory usage: 199388 KiB
> Rank 61 reached 100 vertices from seed 69, flow 1
> Speed RAY_SLAVE_MODE_EXTENSION 25 units/second
> Rank 61: assembler memory usage: 199360 KiB
> Rank 12 reached 700 vertices from seed 27, flow 2
> Speed RAY_SLAVE_MODE_EXTENSION 19 units/second
> Rank 12: assembler memory usage: 199128 KiB
> mpiexec: Warning: tasks 0-39,41-63 died with signal 15 (Terminated).
> mpiexec: Warning: task 40 died with signal 11 (Segmentation fault).
>

Which version are you using ?

> Is there something I did wrongfully?

Can you provide the command you used to launch Ray ?


It should be something like:

mpiexec -n 64 Ray -k 21 \
-o Test21 \
-s SRR131208.fastq \
-p SRR131208_1.fastq SRR131208_2.fastq  


The assembly I got:


Contigs >= 100 nt
  Number: 149053
  Total length: 75819372
  Average: 508
  N50: 822
  Median: 319
  Largest: 8701
Contigs >= 500 nt
  Number: 54663
  Total length: 54570619
  Average: 998
  N50: 1068
  Median: 849
  Largest: 8701
Scaffolds >= 100 nt
  Number: 140427
  Total length: 77365612
  Average: 550
  N50: 968
  Median: 290
  Largest: 16974
Scaffolds >= 500 nt
  Number: 46037
  Total length: 56116859
  Average: 1218
  N50: 1466
  Median: 887
  Largest: 16974




> So far I have found Ray less memory demanding and hope
 > to be able to use it. Please let me know if you can help.
>


I think we will work it out.


> Also, would you recommend me to trim or mask low quality sequences from the 
> data
 >(80M reads)? I saw your reply to another user, but did not specify 
whether trimming could help Ray significantly.
>

Better input should yield better assemblies [1], but it is not the 
reason why Ray is throwing an error here.


[1] http://genome.cshlp.org/content/early/2012/01/12/gr.131383.111

> Thanks,
>
>
> WJ



------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Denovoassembler-users mailing list
Denovoassembler-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/denovoassembler-users

Reply via email to