abyss-pe file prob double /1

Question

[Siva]

Hi guys,

I really don't understand, always got a problem to make an abyss with real data :(

Got this in my log (full log here: )

abyss-map: missing arguments
abyss-fixmate: error: All reads are mateless. This can happen when first and second read IDs do not match.
error: `../FQ_file/Cfinal_bwa_refXRQ_SF099_primALN_mapped_PE.fq-3.hist': No such file or directory
make: *** [../FQ_file/Cfinal_bwa_refXRQ_SF099_primALN_mapped_PE.fq-3.dist] Erreur 1
make: *** Destruction du fichier « ../FQ_file/Cfinal_bwa_refXRQ_SF099_primALN_mapped_PE.fq-3.dist »

I got this version of Abyss: 1.5.2 and launch this command:

abyss-pe contigs mp="" clean name=mappedR_SEPE k=60 j=25 s=250 in='PE.fq' pe='PE.fq' se='SE.fq'

Actually in my PE.fq file there is /1 and /2 paires, but sometimes two /1 can be retrieved, and then got an interleaved file. But in manuals, it says it can work even with that format.

example (PE.fq):

advance thanks

[Siva]

I got these files, but not final contig file:

coverage.hist
mappedR_XRQrefSF099_SEPE-bubbles.fa
mappedR_XRQrefSF099_SEPE-1.fa
mappedR_XRQrefSF099_SEPE-1.adj
mappedR_XRQrefSF099_SEPE-2.adj
mappedR_XRQrefSF099_SEPE-1.path
mappedR_XRQrefSF099_SEPE-3.adj
mappedR_XRQrefSF099_SEPE-2.path
mappedR_XRQrefSF099_SEPE-3.fa
mappedR_XRQrefSF099_SEPE-indel.fa

Answer 1

[benv]

Hi Siva,

Try this instead:

$ abyss-pe contigs clean name=mappedR_SEPE k=60 j=25 s=250 in='PE.fq'

The fact that you have duplicate reads with the "/1" suffix in your file indicates that there is a serious problem with your data. In order to get a good assembly, you will probably need to solve that problem.

[Siva]

Hi Benv, I'll not try with PE only because I wanted to make contigs with single end reads also.

Then yea I also think that problem is coming from my files, I got it from a "bwa mem" mapping and so have none unique alignment that's why having multiple /1 or multiple /2.

But I've made a filtering before with -F 256 with samtools, so don"t understand why there is yet not primary alignment.

I'll check...