Question: Finding the right GO annotation flat file for human build GCRh37
0
gravatar for summerela
4.3 years ago by
summerela110
United States
summerela110 wrote:

I found a download for the latest version of the GO annotation flat file here: 

http://geneontology.org/gene-associations/gene_association.goa_ref_human.gz

But I'm also trying to locate a version that goes with the GRCh37 build, and I cannot figure out how to tell which archived GO database to use. I have not heard back from their support.

Is it safe to assume that a GO database release in February 2014 corresponds to Ensembl 75/GRCh37 since they are both released on the same date? 

I would really appreciate it if anyone can point me in the right direction. Thanks! 

rna-seq go • 1.2k views
ADD COMMENTlink modified 4.3 years ago by Jean-Karim Heriche20k • written 4.3 years ago by summerela110

what's the specific link between GOA and GRCh37 ?

ADD REPLYlink written 4.3 years ago by Pierre Lindenbaum121k

I see what you mean there!  

I guess more specifically, I need to figure out which version of Uniprot was used? I'm new to working with a species that actually has a reference so please excuse my newbie-ness; if they mapped the GO terms to a version of uniprot based on GRCh37 vs a version of uniprot based on GRCh38, wouldn't that make some positional differences? 

ADD REPLYlink written 4.3 years ago by summerela110
2
gravatar for Jean-Karim Heriche
4.3 years ago by
EMBL Heidelberg, Germany
Jean-Karim Heriche20k wrote:

If you need GO annotations for Ensembl genes, just get them from Ensembl for the specific version you're interested in. You can do so using BioMart or the API. Uniprot-GOA is not concerned with a reference genome, it annotates protein sequences and other resources like Ensembl do the mapping to a reference genome. So if you want to use GO terms associated with some Uniprot IDs and need to map those to a particular reference then just find which genes the Uniprot IDs are associated with in the reference genome you're interested in.

ADD COMMENTlink written 4.3 years ago by Jean-Karim Heriche20k

This all makes complete sense. It doesn't matter the build, because the protein is the final product, no matter what the location on the genome. Thanks so much! 

ADD REPLYlink written 4.3 years ago by summerela110
Please log in to add an answer.

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 711 users visited in the last hour