NNN in maker output transcript

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

NNN in maker output transcript

Wen Yao
Dear guys,

Recently, I run maker to annotate a genome. I found that the transcript fasta file output by Maker contains "NNN". Is this normal?
If not, what's going on? Is this a bug of maker or my configuration of maker is not correct? 
I told maker to use snap and augustus for de novo prediction and use exonerate to align ESTs and proteins.

Thanks!

Wen Yao

_______________________________________________
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: NNN in maker output transcript

Daniel Ence-2
Hi Wen Yao,

Do you mean that some of the transcript sequences contain “N” characters or that an entire transcript sequence is “NNN”?  



> On Oct 25, 2017, at 3:25 AM, Wen Yao <[hidden email]> wrote:
>
> Dear guys,
>
> Recently, I run maker to annotate a genome. I found that the transcript fasta file output by Maker contains "NNN". Is this normal?
> If not, what's going on? Is this a bug of maker or my configuration of maker is not correct?
> I told maker to use snap and augustus for de novo prediction and use exonerate to align ESTs and proteins.
>
> Thanks!
>
> Wen Yao
> _______________________________________________
> maker-devel mailing list
> [hidden email]
> 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: NNN in maker output transcript

Carson Holt-2
In reply to this post by Wen Yao
The gene predictor generates the model. I don’t think snap will generate a model that contain an N. Augustus might be able to across a single codon (I’m not sure there). The N means that the nucleotide is unknown (i.e. it can be A, T, C or G). An NNN codon produces the amino acid X (which is the unknown amino acid code). So it is possible that for something as short as one or two codon’s that the predictor thinks it’s ok to assume that it will produce a valid codon and uses it to complete the reading frame. Alternatively if you are using est2genome=1 or est_gff then what you are seeing is just the result of an alignment which can align to a couple of N's. You should not use est2genome=1 for anything but training. Also est_gff or pred_gff will not be filtered if you supplied an feature location that includes an N.

—Carson




> On Oct 25, 2017, at 1:25 AM, Wen Yao <[hidden email]> wrote:
>
> Dear guys,
>
> Recently, I run maker to annotate a genome. I found that the transcript fasta file output by Maker contains "NNN". Is this normal?
> If not, what's going on? Is this a bug of maker or my configuration of maker is not correct?
> I told maker to use snap and augustus for de novo prediction and use exonerate to align ESTs and proteins.
>
> Thanks!
>
> Wen Yao
> _______________________________________________
> maker-devel mailing list
> [hidden email]
> 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: NNN in maker output transcript

Carson Holt-2
Also you can check the source of the model by looking a the name. i.e. does it have, augustus, snap, or est2genome in the name?

—Carson

> On Oct 25, 2017, at 9:42 AM, Carson Holt <[hidden email]> wrote:
>
> The gene predictor generates the model. I don’t think snap will generate a model that contain an N. Augustus might be able to across a single codon (I’m not sure there). The N means that the nucleotide is unknown (i.e. it can be A, T, C or G). An NNN codon produces the amino acid X (which is the unknown amino acid code). So it is possible that for something as short as one or two codon’s that the predictor thinks it’s ok to assume that it will produce a valid codon and uses it to complete the reading frame. Alternatively if you are using est2genome=1 or est_gff then what you are seeing is just the result of an alignment which can align to a couple of N's. You should not use est2genome=1 for anything but training. Also est_gff or pred_gff will not be filtered if you supplied an feature location that includes an N.
>
> —Carson
>
>
>
>
>> On Oct 25, 2017, at 1:25 AM, Wen Yao <[hidden email]> wrote:
>>
>> Dear guys,
>>
>> Recently, I run maker to annotate a genome. I found that the transcript fasta file output by Maker contains "NNN". Is this normal?
>> If not, what's going on? Is this a bug of maker or my configuration of maker is not correct?
>> I told maker to use snap and augustus for de novo prediction and use exonerate to align ESTs and proteins.
>>
>> Thanks!
>>
>> Wen Yao
>> _______________________________________________
>> maker-devel mailing list
>> [hidden email]
>> 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