SOAPdenovo2 segfaults on Arch Linux

[michaelgruenstaeudl]

Dear Holm,

I am running IOGA on a dataset (paired-end Illumina reads) from which I had successfully assembled complete plastid genomes in the past. When running IOGA on the same dataset today, IOGA reports that SOAPdenovo2 fails under every k-mer level evaluated. IOGA then exits with an index error for variable best, which is probably empty:

Iteration 1
[BAR02A_S1_L001.1] BBmap
[BAR02A_S1_L001.1] Sorting SAM files
[BAR02A_S1_L001.1] Removing duplicates
[BAR02A_S1_L001.1] Extracting reads
[BAR02A_S1_L001.1] Writing forward reads
[BAR02A_S1_L001.1] Writing reverse reads
[BAR02A_S1_L001.1] Old size = 0
[BAR02A_S1_L001.1] New size = 3093513
[BAR02A_S1_L001.1] Running SOAPdenovo2 k = 33 -- 10:32:42.149468
[BAR02A_S1_L001.1] SOAPdenovo2 k = 33 failed
[BAR02A_S1_L001.1] Running SOAPdenovo2 k = 37 -- 10:32:42.313427
[BAR02A_S1_L001.1] SOAPdenovo2 k = 37 failed
[BAR02A_S1_L001.1] Running SOAPdenovo2 k = 41 -- 10:32:42.499712
[BAR02A_S1_L001.1] SOAPdenovo2 k = 41 failed
[BAR02A_S1_L001.1] Running SOAPdenovo2 k = 45 -- 10:32:42.672296
[BAR02A_S1_L001.1] SOAPdenovo2 k = 45 failed
[BAR02A_S1_L001.1] Running SOAPdenovo2 k = 49 -- 10:32:42.864960
[BAR02A_S1_L001.1] SOAPdenovo2 k = 49 failed
[BAR02A_S1_L001.1] Running SOAPdenovo2 k = 53 -- 10:32:43.026761
[BAR02A_S1_L001.1] SOAPdenovo2 k = 53 failed
[BAR02A_S1_L001.1] Running SOAPdenovo2 k = 57 -- 10:32:43.212315
[BAR02A_S1_L001.1] SOAPdenovo2 k = 57 failed
[BAR02A_S1_L001.1] Running SOAPdenovo2 k = 61 -- 10:32:43.401318
[BAR02A_S1_L001.1] SOAPdenovo2 k = 61 failed
[BAR02A_S1_L001.1] Running SOAPdenovo2 k = 65 -- 10:32:43.580864
[BAR02A_S1_L001.1] SOAPdenovo2 k = 65 failed
[BAR02A_S1_L001.1] Running SOAPdenovo2 k = 69 -- 10:32:43.760832
[BAR02A_S1_L001.1] SOAPdenovo2 k = 69 failed
[BAR02A_S1_L001.1] Running SOAPdenovo2 k = 73 -- 10:32:43.963374
[BAR02A_S1_L001.1] SOAPdenovo2 k = 73 failed
[BAR02A_S1_L001.1] Running SOAPdenovo2 k = 77 -- 10:32:44.130588
[BAR02A_S1_L001.1] SOAPdenovo2 k = 77 failed
[BAR02A_S1_L001.1] Running SOAPdenovo2 k = 81 -- 10:32:44.322826
[BAR02A_S1_L001.1] SOAPdenovo2 k = 81 failed
[BAR02A_S1_L001.1] Running SOAPdenovo2 k = 85 -- 10:32:44.507801
[BAR02A_S1_L001.1] SOAPdenovo2 k = 85 failed
[BAR02A_S1_L001.1] Running SOAPdenovo2 k = 89 -- 10:32:44.688810
[BAR02A_S1_L001.1] SOAPdenovo2 k = 89 failed
[BAR02A_S1_L001.1] Running SOAPdenovo2 k = 93 -- 10:32:44.867438
[BAR02A_S1_L001.1] SOAPdenovo2 k = 93 failed
[BAR02A_S1_L001.1] Running SOAPdenovo2 k = 97 -- 10:32:45.051394
[BAR02A_S1_L001.1] SOAPdenovo2 k = 97 failed
Traceback (most recent call last):
  File "/home/michael_science/git/IOGA//IOGA.py", line 432, in <module>
    main(args.reference,args.name,args.forward,args.reverse,args.threads,args.insertsize,args.maxrounds,args.verbose)
  File "/home/michael_science/git/IOGA//IOGA.py", line 377, in main
    source,FP,RP,final_iteration = IOGA_loop(name,ref,forward,reverse,insertsize,threads,maxrounds)
  File "/home/michael_science/git/IOGA//IOGA.py", line 359, in IOGA_loop
    best,k = run_soapdenovo(folder,prefix,FP,RP,insertsize,threads)
  File "/home/michael_science/git/IOGA//IOGA.py", line 254, in run_soapdenovo
    best=sorted(n50,key = lambda x: int(x[1]),reverse = True)[0][0]
IndexError: list index out of range

I strongly suspect that a recent Python update (or an update of one of the Python dependency packages) has corrupted the IOGA code such that the wrapper function for SOAPdenovo2 (i.e., run_soapdenovo()) no longer interacts with the assembler properly. Why else would SOAPdenovo2 fail under every k-mer level on a dataset where an assembly worked nicely before.

Edit 1:
I found that each of the SOAPdenovo2 subfolders generated by IOGA during the assembly process (i.e., BAR02A_S1_L001.1.soap_33, BAR02A_S1_L001.1.soap_37, ..., BAR02A_S1_L001.1.soap_97) contains merely two files: an empty log file and an error file with the following contents:

Version 2.04: released on July 13th, 2012
Compile Jul  9 2013	11:57:30

Evidently, SOAPdenovo2 is not called properly.

Can you please check the code?!

[holmrenser]

Thanks for the error report. I'll dive into the issue ASAP, in the meantime, can you provide me with the following information?

• python version used
• whether you are using SOAPdenovo as installed through IOGA, or whether you have your own installation.

[michaelgruenstaeudl]

• I am running IOGA with Python2.7
• IOGA is using SOAPdenovo as installed through your setup script setup_IOGA.py (i.e., it is running IOGA/exe/SOAPdenovo2-bin-LINUX-generic-r240/SOAPdenovo-127mer)

[michaelgruenstaeudl]

Upon closer examination, I found that the SOAPdenovo version that is installed through your setup script cannot be run on my system:

~/git/IOGA/exe/SOAPdenovo2-bin-LINUX-generic-r240/SOAPdenovo-127mer all

Version 2.04: released on July 13th, 2012
Compile Jul  9 2013	11:57:30
Segmentation fault (core dumped)


~/git/IOGA/exe/SOAPdenovo2-bin-LINUX-generic-r240/SOAPdenovo-127mer pregraph

Version 2.04: released on July 13th, 2012
Compile Jul  9 2013	11:57:30
Segmentation fault (core dumped)

Here are my kernel details:

Linux 4.15.11-1-ARCH
x86_64 GNU/Linux

[holmrenser]

Given that the error you get from the SOAP binary is equivalent to what you see in the error file I am not sure this is a problem with IOGA? It seems like somehow your system stopped being able to run the same SOAP binary? When did you install IOGA/SOAP and have you tried reinstalling it?

Also, it seems like you are running Arch-linux, have you been successful in the past with running IOGA (including this SOAP binary) on Arch?

[michaelgruenstaeudl]

Hi Holm,
yes, I have successfully run IOGA on Arch Linux in the past. In fact, the last time I ran IOGA (Nov 2017), everything worked without fail. Thus, I agree that my issue is not with IOGA itself, but with the SOAPdenovo2 binary. I will close the issue here and rather continue at: aquaskyline/SOAPdenovo2#39