Identifying Proteins (In A Disease) And Protein Targets (For Drugs)

Question

[Shweta]

How can I find proteins involved in a disease (I'm interested in epilepsy) and protein targets (for drugs) associated with it?

Answer 1

[Lars Juhl Jensen]

With the danger of sounding snarky: did you try Google? Less than five minutes got me this far.

Genetics of epilepsy:

• http://www.carpedb.ua.edu/
• http://www.epigad.org

Epilepsy drugs:

• http://www.epilepsy.org.uk/info/treatment/uk-anti-epileptic-drugs-list
• http://www.epilepsy.com/EPILEPSY/SEIZURE_MEDICINES

All you would have to do to find know drug targets is to look up the drugs from these lists in any of a number of general drug-target databases.

Answer 2

[Khader Shameer]

Nice suggestions by Lars and Lyco:

Here is what I would do:

Instead of using a single database, I will use three genetic disease resources and one generic disease resource and go for a union / intersection of the four database (depending on the stringency and class of epileps diseases you need for your analysis).

Genetic disease resource:

• OMIM: you can download data and parse it for Epilepsy.
• Genetic Association Database: You can find genes associated with different classes of epilepsy here
• HUGENavigator: I got 19 disease for epilepsy via Phenopedia search and 26 genes for epilepsy via Genopedia and 6 GWAS hits

Generic disease annotation resource:

• Disease Ontology annotation:

  Click here for annotations from OMIM, GAdb and DO Lite annotations.

Along with target database suggested by Lars and Lyco, you can do some literature curation to see how many of them are actual drug targets. You may also check these literature curated resources for finding targets among your list of proteins involved in epilepsy.

• DrugBank
• Therapeutic Target Database
• MATADOR

Answer 3

[Lyco]

Lars already pointed towards some epilepsy-specific resources. More generally, there are databases and search services that provide links between genes/proteins and disease conditions. Here are two that I use a lot: Fable and GoGene/GoDisease. Both work by pubmed mining, i.e. the genes that are returned are not necessarily causally involved in the disease but are often cited in this context. For most applications, this is what you need (and there is a big overlap with the causal genes). If you are really into causal genes, you should have a look at OMIM.