What are novel genes?

**higherme** · Oct31-09, 01:38 AM

Hi
I don't know what novel genes are. Any help?

Thanks

**Monique** · Oct31-09, 06:24 AM

Those are pieces of DNA that had not previously been identified as being a gene. How did you come to this question?

**higherme** · Nov1-09, 12:23 AM

because i'm trying to find orthologs on this gene called san-1 from C. elegans, But it doesn't seem to have any. So my assignment saids to try finding it with its novel gene.

So novel gene is not really a gene? I'm confused. does it have a name though?

**qraal** · Nov1-09, 02:37 AM

Originally Posted by higherme

because i'm trying to find orthologs on this gene called san-1 from C. elegans, But it doesn't seem to have any. So my assignment saids to try finding it with its novel gene.

So novel gene is not really a gene? I'm confused. does it have a name though?

Isn't a gene that doesn't belong to a previous identified gene family, thus it has no comparable genes in other species?

**qraal** · Nov1-09, 02:42 AM

Originally Posted by qraal

Isn't a gene that doesn't belong to a previous identified gene family, thus it has no comparable genes in other species?

Apparently not. According to this source... Ensembl Human genome release a 'novel gene' is...

Novel genes are genes that have been predicted by Ensembl on the basis of similarity to protein or cDNA sequences. They cannot be mapped with confidence to existing entries in any public sequence database.

...so kind of what I said and kind of not.

**Monique** · Nov1-09, 06:11 AM

Novel means new, you can say those are uncharacterized genes.

I don't understand that you can't find the ortholog of san-1. In which databases did you look?

**higherme** · Nov1-09, 02:07 PM

I used NCBI Blast and Ensembl.
Blast results gave Mad3/Bub1 domains when searched.. but i don't think those are orthologs? they are just a domain that san-1 has??
and ensembl didn't find any.

**Monique** · Nov1-09, 02:34 PM

It gave domains when you did a blast, did you try blasting the protein sequence to other species? That's how you find orthologs. The fact that it has a Mad3 domain is a big clue as to the function and possible orholog of the protein.

Besides doing the interspecies blast search you can also search the literature on Pubmed and the C. elegans worm database at Wormbase.

**higherme** · Nov1-09, 11:14 PM

Originally Posted by Monique

It gave domains when you did a blast, did you try blasting the protein sequence to other species? That's how you find orthologs. The fact that it has a Mad3 domain is a big clue as to the function and possible orholog of the protein.

Besides doing the interspecies blast search you can also search the literature on Pubmed and the C. elegans worm database at Wormbase.

yep, i used Blastp against other species. There were some "hypothetical proteins".
other than that, the % identity is pretty low (less than 25%)

**Monique** · Nov3-09, 09:06 AM

It's best to first blast against species not too divergent from the one you are working with. Going from a nematode to humans can be a big evolutionary leap. If you do a Blastp against the Drosophila melanogaster species you should get Bub1-related kinase as a top hit.

For divergent sequences it is best to use a PSI-blast, it goes through several iterations and can therefore find a better match. Doing an alignment is only one step to find possible orthologs, you should next check the actual data on the function of the SAN-1 protein and its possible ortholog.