Genome Annotation tips

[Ashu]

Can any one give me some tips about genome anotation for big genome, which tools or pipelines are good
Thank you in advance,

[Richard Finney]

You might want to give more details on your "big genome". Large genomic data? new species? what species? De-novo assembled?

[FredericRaymond]

General advice: Read the supplementary material from published genome papers, compare their methods and find which one would be better for you project.

[Ashu]

Thank you both for your reaction :-)
@Richard, it's size is 3 GB, de novo assembled from short read. I am trying to follow the ensemble gene build pipeline, but it is not very clear. I was just wondering how other people do it. I have also seen in literature that people follow the ensemble pipeline. If there is any other tip, that will be nice
@FredericRaymond, thank you, I have done that already, but it is very vague.
Thank you again
ashu

[boetsie]

one program I know of is AUGUSTUS, but i've never used it myself

[FredericRaymond]

I've used Augustus for the annotation of the Leishmania tarentolae genome. The software is built for prokaryotes, but we collaborated with the creator of the software to adapt it for Leishmania, which do not have introns. It performed pretty well.

[boetsie]

Ehhh, i see a contradiction in your reply FredericRaymond, i think you meant that the software is for eukaryotes. The website (http://augustus.gobics.de/) says

AUGUSTUS is a program that predicts genes in eukaryotic genomic sequences

[Ashu]

Thank you all for the tip, Augustus was in my list, but I was focused on the ensemble Pipeline. I will give it a try and see if it is a good tool.
Thank you again,
cheers
Ashu

[FredericRaymond]

You right boetsie. My mistake. However, the intron-less models can be built for Augustus.

FR

[bbsinfo]

how do you guys find possible introns on a gene? Which software do you use?

[yumtaoist]

Genewise is used very frequently

[Hobbe]

bbsinfo, there are (as far as I know) three ways to find introns.

1) You could do a spliced alignment of some kind of assembled transcripts/ESTs using for example Blat (if you are working in nucleotide space that is)
2) Get junctions from mapped transcripts using for example Tophat
3) Align known proteins to the genome using Genewise, Exonerate or perhaps Scipio.

Check out the suggested pipeline on the PASA homepage, it works well:

http://pasa.sourceforge.net/

[gprakhar]

Thank you all for the tip, Augustus was in my list, but I was focused on the ensemble Pipeline. I will give it a try and see if it is a good tool.
Thank you again,
cheers
Ashu

Hello,

Did you use the Ensemble pipeline?
If yes can you provide a few details!

Regards,
--
pg