That Sevens database looks great, but why-oh-why do they hide their data in an imagemap?! (e.g., http://sevens.cbrc.jp/detail.php?db=hsap&chr=1&l=23073612&r=25074694&twin=0&id=CBRC-HSAP-01-0032)
I have a list of GPCRs, and I'd like to determine the G-proteins (Gs, Gi, Gq) to which they bind. In my ideal world, this would be a simple database lookup of experimentally-determined results. But in my searching, I've not found such a database.
Alternatively, I'd be open to computational prediction of G-protein specificity. For example, I've found these two publications:
- GRIFFIN: a system for predicting GPCR-G-protein coupling selectivity using a support vector machine and a hidden Markov model.
- Prediction of GPCR-G protein coupling specificity using features of sequences and biological functions.
But it doesn't appear like either of these methods are widely used. Anyone else know of any resources/methods that I should be aware of?
3 answers
You can try to classify your sequences using GPCRpred to identify their family or sub-family information. A webserver is available based on the method. List of related resources are provided in the information page of GPCRpred. If you have large number of sequences, you may request the authors for a Mechanize based access to the server.
If you need a non-prediction based approach, I would recommend you to take a look at the Sevens database or GPCRDB.
Just added GPCRDB to the answer, they provide web-services (http://www.gpcr.org/7tm/webservices;jsessionid=07269f5e3769610b6dd012d36dc2) and BLAST searches.
[?]
http://bioinformatics.biol.uoa.gr/gpDB/ [?]You can also check the additional resources (databases, web servers) provided at the following webpage :[?] http://www.bioinformatics.fr/biology.php?subsection=GPCR
the interface of gpDB is a bit crude, but the data there is very relevant and useful. Doesn't appear to be an API or download files, but at least I could screen scape if necessary. Thanks!
Log in to answer this question.