Extracting genes from newly sequenced data

Discussions about predicting genes with AUGUSTUS. Not covered here: WebAUGUSTUS and BRAKER1

Moderator: bioinf

Post Reply
Bharat
Posts: 1
Joined: Fri Mar 11, 2016 3:26 am

Extracting genes from newly sequenced data

Post by Bharat »

Hi,

I have a file containing contigs from a newly sequence fungi in my lab. I am interested in finding the genes from the contigs. I saw that I need to train the data but the problem is that I don't know which fungal species to choose for training ?? Any help would be highly appreciated.
User avatar
katharina
Site Admin
Posts: 531
Joined: Wed Nov 18, 2015 6:14 pm
Location: Greifswald
Contact:

Re: Extracting genes from newly sequenced data

Post by katharina »

You have 2 options:

1) Use parameters of a related species for which AUGUSTUS has already been trained to predict genes in your novel sequences. Check which trained species is taxonomically closest to your species.

2) Train AUGUSTUS for your target species. This requires a high quality genome assembly, and either substancial cDNA data, protein sequence data or RNA-Seq data. For the first two, try WebAUGUSTUS, for the last, try BRAKER1.
gclongo
Posts: 1
Joined: Sat Apr 16, 2016 7:45 pm

Re: Extracting genes from newly sequenced data

Post by gclongo »

Hi,

I also have a novel genome that is need of gene predictions. It is a teleost fish and the closest relative I see is the zebra danio. Are there specific options I should use when running Augustus that informs the program the genome is a relative, not the exact taxa? If this requires specific flags when using the closest relative, is it necessary to run it on my server to invoke these flags or is it ok to upload it to the Augustus server? On another note, do you guys have any plans for getting a cichlid gene prediction up and running any time soon?

Thanks for your time!

Cheers,
Gary
User avatar
katharina
Site Admin
Posts: 531
Joined: Wed Nov 18, 2015 6:14 pm
Location: Greifswald
Contact:

Re: Extracting genes from newly sequenced data

Post by katharina »

There are no special flags involved. AUGUSTUS will not be aware of how well a certain parameter set "fits" to a particular genome.

I am currently not aware of a cichlid gene prediction project from our side.
Post Reply