IDBA-UD Invalid insert distance and Segmentation Fault

[swapnil shevate]

Hello,

I get this error segmentation fault every time I run IDBA.

I have looked at all the old threads related and I found that -

a. Its an version issue that Illumina's latest version doesnt support older IDBA but I am using the latest IDBA.

Still the same error -   Segmentation Fault

number of threads 80

reads 200770

long reads 0

extra reads 0

read_length 108

kmer 20

kmers 4919995 4894071

merge bubble 4143

contigs: 11948 n50: 346 max: 3736 mean: 242 total length: 2894301 n80: 231

aligned 61742 reads

confirmed bases: 537493 correct reads: 2076 bases: 266

distance mean -80 sd 0

invalid insert distance

kmer 40

kmers 2622328 2611678

merge bubble 14

contigs: 5724 n50: 418 max: 3736 mean: 406 total length: 2325625 n80: 293

aligned 59472 reads

confirmed bases: 532696 correct reads: 2308 bases: 60

distance mean -80 sd 0

invalid insert distance

kmer 60

kmers 2004924 1999221

merge bubble 1

contigs: 4846 n50: 441 max: 3736 mean: 440 total length: 2135890 n80: 316

aligned 56055 reads

confirmed bases: 512318 correct reads: 2356 bases: 10

distance mean -80 sd 0

invalid insert distance

kmer 80

kmers 1754871 1750032

merge bubble 0

contigs: 4214 n50: 466 max: 3736 mean: 467 total length: 1969978 n80: 337

aligned 52660 reads

confirmed bases: 488738 correct reads: 2327 bases: 0

distance mean -80 sd 0

invalid insert distance

segmentation fault

b. Next i found a thread that says using --num_threads 1 eliminates the problem.

But I still get the same error.

command used - idba_ud -r 05012013_ACAAAC_L001_R1_R2_NEW.fasta -o output

My input file - .fasta

>D4ZHLFP1:55:D29RLACXX:1:1113:17167:23673 1:N:0:ACAAAC

ATGATGCCAACGTCCAACTGAGCGAGCCACATCAGTACCGCCAGACCGAGCAGCATGGGGCTGCCCAGCGAGGCACGGCGCCCCGTCAGCTGGGCGCCGCCGAGCAGG

>D4ZHLFP1:55:D29RLACXX:1:1113:17187:23691 1:N:0:ACAAAC

AATATTTTTTGAACAGTTGCTAAACCTGAACGCTTTTTGCTTGGTAAATCGATTTGAGTAATATCACCAGTTATTACGGCTTTTGAGTTAAAGCCAATCCTGGTCAAA

>D4ZHLFP1:55:D29RLACXX:1:1113:17118:23707 1:N:0:ACAAAC

GGCGACACTGCGTGGACCATCCTTGCCCATGGTCTGTTGCATGCGTTTGCGGGTGACCAGAACCTGGATCAGGCTGCTGATGATGTAGCCAAGCACGAACGCCCAGAG

>D4ZHLFP1:55:D29RLACXX:1:1113:17010:23710 1:N:0:ACAAAC

CAAGGAATACCAGATCATGCGCGACGCATCGCTTGCGGTGCTGCGCGAGATCGGCGTCGAGACCGGCGGCTCCAACGTGCAGTTTGGCGTCGACCCGGATACCGGACG

>D4ZHLFP1:55:D29RLACXX:1:2302:9821:43622 1:N:0:ACAAAC

TTTAAAGTCTGGCCAATTGTATATTCAGCCTATTTAAGTCTGCATGAAATGAGCGGCTTAAATACGTCTAATTTTGTTGGTCTTCAAAATTATTTTGAAGTTTTTTCT

>D4ZHLFP1:55:D29RLACXX:1:2302:9807:43733 1:N:0:ACAAAC

I am not sure if this is a problem with the input file ?

Is there any solution for this.

Please help me with this if anyone knows whats the issue here ?

Sincerely,

[Stephen Turner]

I and several others have confirmed this segfault on another post: https://groups.google.com/d/msg/hku-idba/RzTkrVTod8o/UWAPxhKQKSEJ

I'm not sure if the author has responded to any of these issues.

[swapnil shevate]

No the posts are still there -

I have looked at the entire discussion and I found just two suggestions the once posted.

I guess someone should definitely have an answer.

So what was your solution for this ? Did you try running it again ?

Any suggestions should be of great help !

[Stephen Turner]

No, I never got this resolved. I think someone mentioned an old version of IDBA worked, but I never tried that.

--
You received this message because you are subscribed to the Google Groups "hku-idba" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hku-idba+u...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[tce...@staffmail.ed.ac.uk]

Version 1.0.9 seems to not have this problem.
I found it here 

http://i.cs.hku.hk/~alse/hkubrg/login.html

[Sunit Jain]

Just wondering, was this issue ever resolved?

[Mark Davids]

You will get these errors for single end reads or not wel formatted fasta for paired reads.
For paired use the supplied script to merge 2 seperate files. Format should be

>r1f
....
>r1r
....
>r2f
...
>r2r
...

For single end reads I use the last contig-x.fa because due to the error the contig.fa might not be fully written.

[Jeffery Hughey]

I couldn't find an answer to the above in any of the newer postings.  I've been playing with IDBA_UD 1.1.0 for about 4 months now.  I ran across this thread regarding invalid insert distance, which I too was experiencing.  I kept getting this error.  I tried to download the old version of IDBA_UD as suggested by many, but it was taken down from the website (https://i.cs.hku.hk/~alse/hkubrg/projects/idba_ud/index.html).  I then noticed that someone above or elsewhere said to convert from fastq to fasta prior to using IDBA_UD.  I did this with my data using the fq2fa command in the bin folder.  Success!!!  The program is running smoothly now.