Question: What'S The Best Method For Assigning Go Slim Terms To Non-Model Plant Genes?
gravatar for jli99
8.7 years ago by
jli99150 wrote:

I want to assign GO slim terms to some plant genes in non-model plants. So far as I know the only plant genomes relatively accurately annotated with GO slims are rice and Arabidopsis. So I am thinking of using best blast hit in rice or Arabidopsis (depends on whether the genome under study is monocot or dicot) to assign the GO slims. I'm working on genome scale dataset so cannot do manual curations, but I want be set up a reasonably accurate procedure. I'm new to this area and would appreciate if somebody can let me know if there are better ways to do this, or things other than blast similarity that I should take into account.

gene function • 5.0k views
ADD COMMENTlink modified 8.5 years ago by Rph10 • written 8.7 years ago by jli99150
gravatar for 2184687-1231-83-
8.7 years ago by
2184687-1231-83-5.0k wrote:

As you mention, you want to infer isofunctionality from homology, you you can use the homology hits to transfer rice and Arabidopsis annotations to your new species. An important caveat here is that if the gene has duplicated after speciacion, which is common in plants, your isofunctionality prediction may be less accurate, since plants tend to sub- and neofunctionalize their gene repertoires after duplication. So try to annotate unique one-to-one hits separately from one-top-many hits.

ADD COMMENTlink written 8.7 years ago by 2184687-1231-83-5.0k

Thanks avilella, this is a very good point indeed.

ADD REPLYlink written 8.7 years ago by jli99150
gravatar for Dhl
8.7 years ago by
Dhl20 wrote:

Hi Jingping, you could try the following two methods: This method assigns function by blast search.

With this method, you first do a genome-wide InterPro domain scan, then assign GO terms based on InterPro2GO mapping.

One additional method is to first build family trees (e.g.; ), then transfer annotations within the protein family. However this method involves manual curation so it might not work for you.

DHL On behalf of the Gene Ontology (GO) Help Desk

ADD COMMENTlink modified 10 months ago by RamRS30k • written 8.7 years ago by Dhl20

Thanks DHL, I appreciate it. InterPro2GO looks more or less unbiased to me considering avilella's point. And I think this would probably make more sense than trying to map to A.thaliana or rice especially for the species specific stuff.

ADD REPLYlink written 8.7 years ago by jli99150

You need to be very careful with the direct mapping of GO terms from homologs. IMHO this may not be an ideal way. For example if the best hit with respect to sequence similarity of a query sequence may not be the correct protein to be used for annotation transfer since paralogous protein sequences from the same organism do share high identity but function may vary. We are developing an approach called SEQ2GO: you can see the abstract that we recently presented at Moscow Conference on Computational Molecular Biology here:

ADD REPLYlink written 8.6 years ago by Khader Shameer18k
gravatar for Rph
8.7 years ago by
Rph10 wrote:

Hi Jingping,

You might also be interested in the electronic GO annotations that are produced by the projection of manual annotations from Arabidopsis and rice to other plant species using Ensembl Compara ortholog data. The species that have been annotated using this method include poplar, grape, brachypodium, maize etc. and there are currently almost 270,000 annotations for these species.

The EBI's QuickGO browser allows you to view these annotations using this link;

and also you will be able to use QuickGO (home page; to view these annotations in a GO slim, either a pre-defined one or one you make yourself.

The annotation method used to create these annotations is GO_REF:0000035, which is described on this page;

If you need any help in using QuickGO to do your slim, please contact

Rachael on behalf of the UniProt-GO annotation project.

ADD COMMENTlink modified 12 months ago by RamRS30k • written 8.7 years ago by Rph10
gravatar for boczniak767
8.7 years ago by
boczniak767690 wrote:

You could try inparanoid, I remember that in addition to data it provides software.

ADD COMMENTlink written 8.7 years ago by boczniak767690

I'm a little cautious about this: does sequence similarity or orthology support more functional similarity?

ADD REPLYlink written 8.7 years ago by jli99150
Please log in to add an answer.


Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 1251 users visited in the last hour