merged tandem duplications

classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|

merged tandem duplications

Yannick Wurm
Hi Carson et al,

thanks for keeping up the good work.
One gene family I'm looking at has some tandem duplications. And Maker
sometimes merges tandem duplicates into a single gene. Probably based
on some of the BLASTX data. Is there an easy solution for this?

(the screenshot I'm providing represents 4 separate genes, not a
single big one.)
http://fourmidable.unil.ch/temp/makerMerging.png

This was run using 2.07beta. But I've seen the same behavior in older
versions.

Thanks,
yannick

_______________________________________________
maker-devel mailing list
[hidden email]
http://box290.bluehost.com/mailman/listinfo/maker-devel_yandell-lab.org
Reply | Threaded
Open this post in threaded view
|

Re: merged tandem duplications

Carson Hinton Holt
Re: [maker-devel] merged tandem duplications There is not an easy solution, although there is a possible way to deal with this in the code.  What MAKER currently does is provide hints on EST confirmed introns and protein suggested coding ORFs.  Prediction algorithms like Augustus and SNAP then decide on gene boundaries themselves.  The way I would deal with this is to provide a hint to SNAP and Augustus to terminate any models and not allow introns to cross specific coordinates.  These programs can then follow or ignore the termination hint as they choose.  I would have to devise a means to use the repeat pattern in blastx evidence to produce the break point hints.

Thanks,
Carson


On 12/15/10 1:47 AM, "Yannick Wurm" <yannick.wurm@...> wrote:

Hi Carson et al,

thanks for keeping up the good work.
One gene family I'm looking at has some tandem duplications. And Maker
sometimes merges tandem duplicates into a single gene. Probably based
on some of the BLASTX data. Is there an easy solution for this?

(the screenshot I'm providing represents 4 separate genes, not a
single big one.)
http://fourmidable.unil.ch/temp/makerMerging.png

This was run using 2.07beta. But I've seen the same behavior in older
versions.

Thanks,
yannick

_______________________________________________
maker-devel mailing list
maker-devel@...
http://box290.bluehost.com/mailman/listinfo/maker-devel_yandell-lab.org



_______________________________________________
maker-devel mailing list
[hidden email]
http://box290.bluehost.com/mailman/listinfo/maker-devel_yandell-lab.org
Reply | Threaded
Open this post in threaded view
|

Re: merged tandem duplications

Yannick Wurm
Thanks for the quick reply.
Yeah I don't know how big a problem it is in other genomes, but for the ant genomes I've seen this a few times. 

Apologies for adding yet another item onto your todo list...
best,
yannick

On 16 Dec 2010, at 08:58, Carson Holt wrote:

There is not an easy solution, although there is a possible way to deal with this in the code.  What MAKER currently does is provide hints on EST confirmed introns and protein suggested coding ORFs.  Prediction algorithms like Augustus and SNAP then decide on gene boundaries themselves.  The way I would deal with this is to provide a hint to SNAP and Augustus to terminate any models and not allow introns to cross specific coordinates.  These programs can then follow or ignore the termination hint as they choose.  I would have to devise a means to use the repeat pattern in blastx evidence to produce the break point hints.

Thanks,
Carson


On 12/15/10 1:47 AM, "Yannick Wurm" <<a href="x-msg://410/yannick.wurm@unil.ch">yannick.wurm@...> wrote:

Hi Carson et al,

thanks for keeping up the good work.
One gene family I'm looking at has some tandem duplications. And Maker
sometimes merges tandem duplicates into a single gene. Probably based
on some of the BLASTX data. Is there an easy solution for this?

(the screenshot I'm providing represents 4 separate genes, not a
single big one.)
http://fourmidable.unil.ch/temp/makerMerging.png

This was run using 2.07beta. But I've seen the same behavior in older
versions.

Thanks,
yannick

_______________________________________________
maker-devel mailing list
<a href="x-msg://410/maker-devel@box290.bluehost.com">maker-devel@...
http://box290.bluehost.com/mailman/listinfo/maker-devel_yandell-lab.org




_______________________________________________
maker-devel mailing list
[hidden email]
http://box290.bluehost.com/mailman/listinfo/maker-devel_yandell-lab.org
Reply | Threaded
Open this post in threaded view
|

Re: merged tandem duplications

Carson Hinton Holt
Re: [maker-devel] merged tandem duplications It happens to a minor degree in most genomes but is not particularly frequent.  When it does happen in organisms with a lower gene density there is less of an issue because they separate out nicely, however it seems to be more frequent in the ant genomes.  I noticed it with the olfactory receptors (one of the targets of manual curation when it we annotated them).  One of the issues with the ants is that short gaps in the assembly overlap the CDS of a gene so programs like exonerate that can normally find the correct boundaries of tandem duplications like this end up stretching the alignment into the neighboring repeat since the correct location is partially covered by NNNNNNNNN.  There also seem to be a few where homopolymers cause sequencing errors and neighboring regions start to align better than the more compact correct alignment.

Thanks,
Carson

On 12/15/10 7:42 PM, "Yannick Wurm" <yannick.wurm@...> wrote:

Thanks for the quick reply.
Yeah I don't know how big a problem it is in other genomes, but for the ant genomes I've seen this a few times.

Apologies for adding yet another item onto your todo list...
best,
yannick

On 16 Dec 2010, at 08:58, Carson Holt wrote:

There is not an easy solution, although there is a possible way to deal with this in the code.  What MAKER currently does is provide hints on EST confirmed introns and protein suggested coding ORFs.  Prediction algorithms like Augustus and SNAP then decide on gene boundaries themselves.  The way I would deal with this is to provide a hint to SNAP and Augustus to terminate any models and not allow introns to cross specific coordinates.  These programs can then follow or ignore the termination hint as they choose.  I would have to devise a means to use the repeat pattern in blastx evidence to produce the break point hints.

Thanks,
Carson


On 12/15/10 1:47 AM, "Yannick Wurm" <yannick.wurm@... <x-msg://410/yannick.wurm@...> > wrote:

Hi Carson et al,

thanks for keeping up the good work.
One gene family I'm looking at has some tandem duplications. And Maker
sometimes merges tandem duplicates into a single gene. Probably based
on some of the BLASTX data. Is there an easy solution for this?

(the screenshot I'm providing represents 4 separate genes, not a
single big one.)
http://fourmidable.unil.ch/temp/makerMerging.png

This was run using 2.07beta. But I've seen the same behavior in older
versions.

Thanks,
yannick

_______________________________________________
maker-devel mailing list
maker-devel@... <x-msg://410/maker-devel@...>
http://box290.bluehost.com/mailman/listinfo/maker-devel_yandell-lab.org





_______________________________________________
maker-devel mailing list
[hidden email]
http://box290.bluehost.com/mailman/listinfo/maker-devel_yandell-lab.org