Molecule
links to relevant pages
Caltech documentation
Example molecule pages
Molecule model build
Contents
Small molecules section of WB2007 Grant
Section 1J. Curate and represent small molecules and physiology
AIM 1J. Small molecules and Physiology
Pharmacological agents have been used extensively to dissect C. elegans behavior and development. For example, Lewis (1980) used levamisole to identify mutations in genes that encode acetylcholine receptor subunits. Horvitz and colleagues used serotonin and imipramine to distinguish classes of egglaying defective mutants (Trent et al., 1983). C. elegans has been used for toxicology studies (e.g. Kampkotter et al., 2007; Sochova et al., 2007) and for drug screens (e.g., Kwok et al., 2006; Voisine et al., 2007). In addition, critical endogenous small molecules that regulate C. elegans have been purified. For example, dafachronic acid (Gerisch et al., 2007; Giroux and Corey, 2007; Motola et al., 2006; Schroeder, 2006) and dauer-inducing pheromones (Butcher et al., 2007; Jeong et al., 2005). This information is also of high value for researchers working on the agricultural and medical nematodes, who seek effective chemical-based controls for these pathogens. The effort associated with this Aim will initially be data modeling and will require minimal curation effort given the current level of activity in this area.
1. Metabolomics.
As described in section 1B, we have initiated collaborations with groups tooling up to generate metabolic profiles for a large panel of C. elegans mutants. The number of researchers undertaking such studies will likely grow in the next five years. These large-scale metabolomic data sets will become an integral part of our effort to comprehensively describe small molecules that are important for the worm physiology.
2. Drug-phenotype.
We will begin the curation process by adding "Small Molecule" to the flags attached to papers screened by Textpresso. Papers flagged in this way will be curated by extracting the names of the compound and the phenotype it causes. We will take advantage of external resources for general information about the compound. For example, we will make a link between serotonin and the corresponding entry in Wikipedia. We will use existing chemical ontologies, such as ChEBI (www.ebi.ac.uk/chebi/) to describe the small molecules referred to in each paper, and will contribute to the ontologies when necessary to add chemical compounds that are absent. To represent these data, we will expand the WormBase data model to create Small_Molecule and Small_Molecule-Phenotype interaction objects, which together describe a small molecule and experiments that report the molecule's contribution to a defined phenotype.
3. Drug-gene interactions (genetic screens).
Once it is established that a small molecule causes a particular phenotype, a genetic screen can identify genes that modify the effects of the small molecule. These genes are thus modifiers of a drug-phenotype interaction. We will represent this three-way connection. We recognize that there can be different biological bases for such interactions, e.g. a mutation could affect drug delivery or metabolism, or even a process parallel to the process that the drug directly affects. We will devise a new database model for small molecules to more formally represents the connections among drugs, genes and phenotypes. Another way drugs can be related to gene sis by expression analysis, e.g., profiling gene expression changes when animals are exposed to a drug. We already have been curating such data by regarding the drug treatment as an experimental condition (e.g. wormbase.org/db/microarray/results?name=WBPaper00026820%3ABnaphthoflavone_1;class=Microarray_experim ent). The small molecule pages will link to expression clusters.
Purpose of model
- to capture chemical and drug entities that have been shown to affect the biology of the worm.
- many biological process (e.g. lifespan, life stage switching, development, neuronal function, etc.) have been assaying such function in the process of chemicals and drugs, we currently capture this information for a couple of different data classes (e.g. gene regulation, phenotype) as a remark as there is no dedicated acedb object class for them.
- to link to other databases that deal with these entities in greater detail.
