MUMmer

Hi Nick,
As I recall, show-tiling only uses sequences directly from the contigs and
does not try and compute any kind of consensus when two contigs overlap, so
I can't think of any reasonable explanation for what you are seeing. If your
dataset is not too big, and you would like me to take a look, please forward
it along.

On another note, I've never been happy with show-tiling's performance and
have been gradually replacing it with the program 'delta-filter'. If you run
'delta-filter -q out.delta > new.delta' you will get a filtered alignment
file that only reports the best matches for each contig to the reference (
i.e. each base of the contig is mapped to one location on the reference).
This program is superior because it allows contigs to be broken and aligned
to different parts of the reference if necessary, but there is no way to
turn this into a pseudo-molecule, yet.

Best,
Adam Phillippy

On 3/14/07, Nick Peters <np1@...> wrote:
>
> Hi,
>
> I'm using show-tiling to create a pseudo-molecule and I'm running into
> problems when trying to mark up the locations of the original contigs on
> it.
>
> Is there a good explanation of how the pseudo-molecule is built
> somewhere or can someone please explain?  There are big chunks of the
> pseudo-molecule that don't perfectly match to any of my original contigs
> so it is obviously doing something rather cleverer than simply laying
> out the contigs in the order of the tiling path padded with Ns.
>
> Apologies if this is documented somewhere I haven't spotted.
>
> Thanks
>
> Nick
>
> -------------------------------------------------------------------------
> Take Surveys. Earn Cash. Influence the Future of IT
> Join SourceForge.net's Techsay panel and you'll get the chance to share
> your
> opinions on IT & business topics through brief surveys-and earn cash
> http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
> _______________________________________________
> MUMmer-help mailing list
> MUMmer-help@...
> https://lists.sourceforge.net/lists/listinfo/mummer-help
>

Hi Amber,
That's a difficult question. Could you expand on your definition of
linearity? Are you looking for the total number of aligned amino acids, or
do you have some definition of an alignment segment in mind? e.g. The number
of alignments that match for over 50aa at 80% idy. There are not any tools
packaged with MUMmer for computing this, but given a simple metric, it would
be easy to write a perl script to compute it based on the alignments.

Best,
Adam Phillippy


On 3/13/07, Amber Hartman <alhartman@...> wrote:
>
> Hi,
>
> I have several promer alignments between somewhat distantly related
> genomes.  I'd like to quantify the amount of "linearity" in each alignment,
> i.e., how well the two genomes have aligned to each other.  I'm a grad
> student in Jonathan Eisen's group and he thought that Steve Salzberg had
> developed some sort of statistic for this a few years ago at TIGR.  I
> couldn't find anything about this in the user manual though.  Does anyone
> over there know anything about this?
>
> Thanks,
>
> Amber Hartman
>
> -------------------------------------------------------------------------
> Take Surveys. Earn Cash. Influence the Future of IT
> Join SourceForge.net's Techsay panel and you'll get the chance to share
> your
> opinions on IT & business topics through brief surveys-and earn cash
> http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
> _______________________________________________
> MUMmer-help mailing list
> MUMmer-help@...
> https://lists.sourceforge.net/lists/listinfo/mummer-help
>
>