[Biodevelopers] Phrap ..parallel version

Ian Korf ik1 at sanger.ac.uk
Sun Feb 23 15:52:04 EST 2003


I haven't looked at the details of the phrap algorithm, but my
understanding is that it makes many calls to cross_match for the
alignments. We sped up RepeatMasker by replacing cross_match with
WU-BLAST, so it may be possible to do the same with phrap. I've meant to
look into this for a while, but it isn't anywhere near the top of my list.
It depended quite a bit on the parameters, but we were able to increase
speed by a factor of 30 without reducing sensitivity. There's a paper on
the subject in Bioinformatics. The name of the program is MaskerAid, and
since it is a generic cross_match bluffer, it might be easy to adapt it to
phrap. But maybe not. If this is a topic of great interest, I could be
convinced to take a closer look. The greatest potential time-sink is
probably not hacking phrap, but doing the experiments to assess the
differences in assemblies.

-Ian

On Sun, 23 Feb 2003, Ravi Kishore wrote:

> Hello all,
> 
>  does anyone know about a parallel version of phrap. The version that i am
> running is a single threaded one and its taking more than a week to finish
> crunching the sequences . If there is a parallel version it would be of
> great help as we have a 16 node cluster.
> 
> Thanks
> 
> ------------------------------------------
> Ravi K Makkena
> # 211-1, Thorn Lane,
> Newark , DE , 19711.
> 
> ph 302 738 7861 (home)
>    302 831 3183 (office)
> 
> __________________________________________
> 
> _______________________________________________
> Biodevelopers mailing list
> Biodevelopers at bioinformatics.org
> https://bioinformatics.org/mailman/listinfo/biodevelopers
> 





More information about the Biodevelopers mailing list