Annotating and finding genes of a non-model organism (no reference genome)

[NGS_New_User]

I desperately need some help in brainstorming how to approach my data. I have a de novo assembled genome of a non-model organism (no available reference genome) and my end goal is to annotate it and find the genes that determine/play a role in sex development. I plan to map/align the assembly to a related species that has an annotated reference genome. The data was sequenced pair end by illumina hiseq.
I am stuck on what to do next, what would be a suggested pipeline on how to go about finding the genes of interest? And what programs/software open source or commercial can I use to achieve that? How do I go about annotating it?
A related post I posted some few weeks ago is below, no responses yet
Please, any ideas/suggestions or directions to an already similar answered post are welcomed.
Thank you very much.

Originally Posted by NGS_New_User

Hello,

I currently ran a de novo assembly on a male and female genome of a non-model organism.
My plan is to compare the male and female genomes (get percentage of similarity and differences) then extract and annotate the sections that are unique between them.
Additionally, I also plan to align(map) both male and female genomes with a well annotated genome of a related species (model organism); and annotate regions of similarities as well as finding out the percentage of how different or similar they are.

My question is, what are the recommended ngs programs (soft ware) that I should use to accomplish what I want to do?

Any suggestions will be appreciated

[rhinoceros]

Predict proteins with FragGeneScan. Blastp against nr. Predict tRNAs with tRNAscan-SE. Predict rRNAs with Blastn against silva SSU and LSURef?

Alternatively, you could consider just submitting your data to IMG (if it's suitable for your organism) and let them do the computations for you? You could at least have a look at their SOP for ideas?

[jimmybee]

It completely depends on your species/sequence.

How big is your genome? Is it repetitive? How close (taxonomically) is the nearest well-annotated genome? Have you outlined functional gene families that you want to target?

Best bet is to identify ORFs with a gene finder (FGENESH, Augustus, GENSCAN), extract sequences and BLAST against a well-annotated gene/protein db of a closely related species...

[Joann]

Originally Posted by NGS_New_User

my end goal is to annotate it and find the genes that determine/play a role in sex development
Thank you very much.

Is this an organism possessing a phenotypic male and female organism in it's reproductive cycle or does it have a more obscure developmental strategy?

[NGS_New_User]

Originally Posted by Joann

Is this an organism possessing a phenotypic male and female organism in it's reproductive cycle or does it have a more obscure developmental strategy?

Yes, it has a phenotypic male and female in its reproductive cycle

[JackieBadger]

Originally Posted by jimmybee

It completely depends on your species/sequence.

How big is your genome? Is it repetitive? How close (taxonomically) is the nearest well-annotated genome? Have you outlined functional gene families that you want to target?

Best bet is to identify ORFs with a gene finder (FGENESH, Augustus, GENSCAN), extract sequences and BLAST against a well-annotated gene/protein db of a closely related species...

Agreed.
Find contigs with ORFs, and then run these through BLAST2GO. You can then filter the annotations based on E-value

[Anemone]

Would like to just follow up on this post a little bit. Is it possible to get some faster annotation than by Blast2GO for non-model organisms? For me at least the blast step is running very slow (meaning weeks), even if using a more powerful computer. Does it help a lot to get a local BLAST database?? Or anyone has experience with the new Blast2GO CLC plugin, does it run faster? Thanks in advance!! :-)

[yueluo]

Originally Posted by Anemone

Would like to just follow up on this post a little bit. Is it possible to get some faster annotation than by Blast2GO for non-model organisms? For me at least the blast step is running very slow (meaning weeks), even if using a more powerful computer. Does it help a lot to get a local BLAST database?? Or anyone has experience with the new Blast2GO CLC plugin, does it run faster? Thanks in advance!! :-)

Setting up a local blast and blast2go database would certaining speed things up if you have access to a server/cluster.