error_corrected2frg segmentation fault

[mindinhead]

Hi there,

I am running assembly using both Illumina and PacBio sequencing data, this assembly run stopped at "error_corrected2frg pe 400 50 2000000000 pe.tmp > pe.linking.frg.tmp" due to the "segmentation fault" it raised. I am using "MaSuRCA-3.2.4" and I did run it successfully many times with other datasets, and this is the first time I had this segmentation fault for "error_corrected2frg". I checked the "pe.tmp" file, it's a complete file and in FASTA format. Since there isn't much information the error message provides, any suggestions would be appreciated. I can share the "pe.tmp" file even though it's a 16G size file... Thank you in advance.

Best, Lin

[olivierarmant]

Dear Lin, I got the same error. Did you succeed to use masurca? In my case I have tried the assembly using PE libraries with 2 different reads length, 250 and 100 bp. The error was specifically with the two libraries with 100bp reads while the 250 run fine. What was your reads length? Olivier

[genaev]

Dear, olivierarmant

In our laboratory, we faced the same problem. We solved this by rewriting the error_corrected2frg original C++ code in python. This works great in our case.

#!/usr/bin/env python

import sys

name_seq = []
ind = -1
with open(sys.argv[5]) as input_file:
    lines = input_file.readlines()
    sys.stdout.write("{"f"VER\nver:2\n"+"}"+"\n""{"
                       f"LIB\nact:A\nacc:{sys.argv[1]}\nori:I\n"
                       f"mea:{sys.argv[2]}\nstd:{sys.argv[3]}\nsrc:\n.\n"
                       f"nft:1\nfea:\n"
                       f"doNotOverlapTrim=1\n.\n""}""\n")
    for line in lines:
        if line.startswith('>'):
            line = line.split(" ")[0]
            line = line[3:]
            name_seq.append(int(line))
            ind += 1
        else:
            sys.stdout.write("{FRG"f"\nact:A\nacc:pa{name_seq[ind]}"
                              f"\nrnd:0\nsta:G\nlib:pa\npla:0\nloc:0"
                              f"\nsrc:\n.\nseq:\n{line}.\nqlt:"
                              f"\n{'E'*(len(line)-1)}\n.\nhps:\n."
                              f"\nclr:0,{len(line)-1}\n""}""\n")
name_seq.sort()
for i in range(0, len(name_seq), 2):
    sys.stdout.write("{"f"LKG\nact:A\nfrg:pa{name_seq[i]}\n"
                      f"frg:pa{name_seq[i+1]}\n""}""\n")

[celinetree]

genaev's work around worked for me to get around this issue!