Header Leaderboard Ad


454 mate pairs and mosaik



No announcement yet.
  • Filter
  • Time
  • Show
Clear All
new posts

  • 454 mate pairs and mosaik

    I am using 454 reads in Mosaik to generate bam files to view alignments with a reference sequence in Artemis (Sanger). Does anyone know a good way to get mate-paired (3kb linker) 454 sff files into Mosaik? Mosaik seems to require separate forward and reverse fasta and qual files (or fastq) for the pairs, which are not easily generated by sff_extract or Newbler.

    Thanks for your advice.


  • #2
    Do you have access to 454 software, specifically newbler (gsAssembler/runAssembly)?


    • #3
      Yes, but the methods I have tried with newbler either convert the 454 mate pair sff to one multi-fasta file with each component read containing a mate pair separated by the adaptor or alternatively one multi-fasta file containing approximately double the number of reads with the adaptor removed. In the latter case, the mate-pairs are present as separate forward and reverse reads but in one multifasta. So I am assuming that I would need to use a script to split this multifasta file into the 2 separate fasta and qual files for the forward and reverse that Mosaik needs. My question is whether I am missing something in newbler or if there is an easier way to do this.


      • #4
        Sticking with Mosaik, it does sound like you'll have to do some data file manipulation. What is your preferred scripting language? It shouldn't be hard using any of the Bio* libraries to split an interleaved FASTA, QUAL or FASTQ file into separate files for forward and reverse reads.

        Alternatively, instead of Mosaik, use the Roche 454 tools to do the alignment directly from the SFF file.


        • #5
          maubp-thank you for the advice. I guess I was thinking perl. The Roche 454 tools have worked fine to do the alignments and generate contigs, but I really wanted to do a check with another method. When I split the reads from the mate-pairs (losing the mate-pair information) or alternatively when I just use the sff files from a normal (non mate-paired) 454 run the graphical view of the Mosaik- and then samtools-generated bam files in Artemis is beautiful and perhaps more immediately informative than newbler.