Description of the project

From GO Wiki
Revision as of 00:22, 15 November 2007 by Erika (talk | contribs)

Jump to: navigation, search

Back to Response to Drug wiki page


In the ChEBI ontology there is "organic molecular compound" CHEBI:25700, defined as a molecular entity that contain carbon and its synonyms are: organic compounds and organic entity. In the GO the definition of "response to organic substance" GO:0010033 term is: a change in state or activity of a cell or an organism (in terms of movement, secretion, enzyme production, gene expression, etc.) as a result of an organic substance stimulus. But, what do we mean for "organic substance"? We can use the definition of ChEBI:25700 for defining what is an organic substance. After that, we can use the "response to organic substance" GO:0010033 as a parent term and moving many terms from "resposne to drug" under this term: see the case of cycloheximidine GO:0046898

When I start to use the ChEBi ontology as a reference to insert new terms in the GO, I relised that I have three cases:
1. the compound in which I am interested is present in the ChEBI molecular structure ontology (example morphine)
Solution: I will stick part or all the ChEBI structure in the GO
2. my compound is not included in the ChEBI molecular structure ontology but it is present in the other ChEBI ontology (application and molecular role) (example: amphetamine)
Solution: First, I will request to ChEBI curators to have the term amphetamine under molecular structure ontology and then I will add a new term in the GO
3. my compound is not included in the ChEBI ontology (example Citalopram) or it is a preliminary entry (example fluoxetine ChEBI 5118)
Solution: I will request a new term to ChEBI people


Cycloximide is a child of resposne to antibiotic (GO:00) but we would like to classify this substance not following its role in the cell but on the base of its moleculare strucuture. So, looking in the ChEBI ontology, we can find that it is a child of cycloalkanes (Fig.1). Using the ChEBI structure, we could implement it in the GO moving the term "response to Cycloheximide" GO 0046898 at the end we have the DAG structure showed in the Fig.1.

Figure 1 [[1]]

We could do the same for response to bacteriocin GO:0046678, response to brefeldin A GO:0031001 and response to streptomycin GO:0046679. For streptomycin it would be easy because it is already present in ChEBI ontology (CHEBI:17076); instead for the others we should request new terms to ChEBI curators and wait until they implement them into the ontology.


"response to morphine" GO:0043278 is an other term having a correspondent one in the ChEBI ontology CHEBI:17303 morphine. The Fig.2 shows the final structure of GO DAG after the implementation of some new terms like "response to isoquinoline alkaloid".

Figure 2 [[2]]


There are many terms of interest to pharmaceutical researchers for the annotation of gene products that responde to the administration of a specific drug, for instance "response to fluoxetine". Fluoxetine is an antidepressant, in detail it is a selective serotonin reuptake inhibitors (SSRIs) which is derived from diphenhydramine, an antihistamine found to inhibit reuptake of the neurotransmitter serotonin (Fig.3)

The list of antidepressants is very long and I have start to request new terrms to CHEBI curators with the aim to add these ones in the GO. In the next CHEBI release, we should have these new terms: SERTRALINE PAROXETINE CITALOPRAM VENLAFAXINE ESCITALOPRAM FLUVOXAMINE BUPROPION DULOXETINE AMITRIPTYLINE DOTHIEPIN

Figure 3 [[3]]

Some Questions and Some Answers:

1) Do we need to stick all the ChEBi structure or can we choose only the terms that we need for classify our GO "response to X" term?
We decided to choose only the terms that we need for insert our "response to X" term in a correct structure
2) Have innumerable children under response to chemical parent term is not a problem. Is it true?
Yes, it is true. We can have as many children as we need.
3) we want to use synonym for better describing the term. For example, "response to fluoxetine" could have a related synonym "response to antidepressant" or a more informative one "response to selective serotonin reuptake inhibitors".Doing that, we could have additional information about the use of a substance as a drug.
Do we all agree with this?
We agree to add synonyms for better describe the use of a substance.
4) fill in the Dbxrefs with the ChEBI ID correspondent to the term in the ChEBI ontology.
5) what about "response to xenobiotic stimulus" GO:0009410? do we want to obsolete it? SF 1494548 on xenobiotic terms
We have to be very careful about that.