Modeling data for Infectious agents
C. elegans serves as an infection model for various bacteria, fungi and nematodes, in order to study:
- host genes involved in infection and resistance
- bacterial virulence factors
- to identify effective drugs
- drug screening for infectious diseases
Model Proposal for ?Infection_assay
?Infection_assay Name ?Text Description ?Text Reagent Bacteria_strain ?Strain Worm_strain ?Strain Symptom Phenotype_observed ?Phenotype Phenotype_not_observed ?Phenotype Dependency Requires_gene ?Gene //genes that make infection happen Requires_molecule ?Molecule //molecules required for infection to happen Requires_treatment Text //certain molecules required for infection to happen Resistence Allele ?Variation //allele that resist infection Drug ?Molecule //molecules that resist infection Treatment Text //treatment that resist infection Regulate_expression Expression_cluster ?Expression_cluster XREF Infection_assay Up_regulate_gene ?Gene #Evidence Down_regulate_gene ?Gene #Evidence Reference ?Paper
- Should add 'Fungal_strain'--kjy (talk) 01:59, 14 February 2014 (UTC)
- 'Viral_strain' as well? -XW
- (Mary Ann) - You should be able to extract whether the Reagent is a fungal/viral/whatever strain by following through from the Strain object to the Species. So, simply adding the Strain name here should be sufficient for the website to display the taxonomic info. Maybe this is preferable to adding lots of new tags?
- I would change 'symptom' to 'Effect_on_nematode', to capture 'lethal', 'slowed_growth' etc. phenotypes. --kjy (talk) 01:59, 14 February 2014 (UTC)
- Although we put Resistant to and susceptibility to pathogen terms in the phenotype ontology, perhaps we should consider moving them out to be organism interaction tags, along with other organism relationship terms such as symbiotic, mutualistic, etc., afterall, shouldn't we have a way to represent Heterorhabditis bacteriophora and its enteric bacterium, Photorhabdus luminescens. I know this is an infection model, but if we go this way, perhaps we should consider opening it up a bit broader to reflect all organism-organism relationships (I guess that also suggests we should change the title of the model to 'Strain_exposure_assay?) --kjy (talk) 01:59, 14 February 2014 (UTC)
- Dependency can go both ways - the gene, molecule, treatment can be required for either susceptibility or resistance. --kjy (talk) 01:59, 14 February 2014 (UTC)
- The Resistance supertag is not quite right. An allele can also confer susceptibility. So perhaps it should be called 'Modifying_influence'; that is, these objects do not define the outcome of the exposure to the possible pathogen, but alter the normal outcome --kjy (talk) 01:59, 14 February 2014 (UTC)
- Some grammar points: 'Regulate_expression' -> 'Regulates_expression', 'Up_regulate_gene' -> 'Up_regulates_gene', 'Down_regulate_gene' -> 'Down_regulates_gene'--kjy (talk) 01:59, 14 February 2014 (UTC)
Model Proposal for ?Gene
The Experimental_model tag refers to ?DO_term (disease), in cases of study of infectious agents using Elegans, it is not so much a 'disease model' as an infection model. It's a bit silly to talk about elegans as a model for dysentery, as compared to saying it is an infection model for the bacteria, Shigella flexneri, which is the pathogen in dysentery.
We will attach this information to Gene, analogous to disease information (similar to the Experimental_model and Potential model tags for Gene). Initially I thought of tagging Genes just as 'Infection_model' but I think it's worthwhile to actually delineate this a bit further and tag genes as either needed for pathogen infection or resistance to it, thus the 'Required_for_infection' and 'Resistance_to_infection' tags.
Current Gene model: ?Gene Disease_info Experimental_model ?DO_term XREF Gene_by_biology ?Species #Evidence Potential_model ?DO_term XREF Gene_by_orthology ?Species #Evidence Disease_relevance ?Text ?Species #Evidence Add under ?Gene: Infection_model Required_for_infection ?Strain XREF Host_gene_required_for_infection ?Species #Evidence Resistance_to_infection ?Strain XREF Host_gene_required_for_resistance ?Species #Evidence Infection_model_description ?Text ?Species #Evidence
Curating information about the infectious agent-Proposal for ?Strain:
As was suggested for capturing information about the bacterial strain/species for GO data, we will use the ?Strain model:
Add to Strain model:
?Strain Infection_model_info Host_gene_required_for_infection ?Gene XREF Required_for_infection Host_gene_required_for_resistance ?Gene XREF Resistance_to_infection Infection_model_description ?Text ?Species #Evidence
Example for Proposal 1
- WBPaper00027729: The Caenorhabditis elegans ABL-1 Tyrosine Kinase Is Required for Shigella flexneri Pathogenesis
In this example the data points are: Gene Required_for_infection = abl-1 (WBGene00000018) The bacterial strain = Shigella flexneri strain 2457T Infection_model_description, is basically a concise description that will describe the model, similar to the 'Disease_relevance' description under Disease_info in ?Gene. Infection_model Required_for_infection ?Strain XREF Host_gene_required_for_infection ?Species #Evidence Resistance_to_infection ?Strain XREF Host_gene_required_for_resistance ?Species #Evidence Infection_model_description ?Text ?Species #Evidence I'd also like to be explicit about the species, as it might be easier for web scripts to pull information if it's right there. But Kevin's point taken that it is a redundant since ?Strain has ?Species information. Modeled in .ace file as: Gene : "WBGene00000018" Required_for_infection "2457T" "Shigella flexneri" Paper_evidence "WBPaper00027729" Required_for_infection "2457T" "Shigella flexneri" Curator_confirmed "WBPerson324" Required_for_infection "2457T" "Shigella flexneri" Date_last_updated "2014-02-12" Infection_model_description "The gram-negative bacterium Shigella flexneri is the causative agent of Shigellosis, a diarrheal disease that affects millions of people every year; the study of Shigellosis in vivo has been hampered by the lack of a suitable model system; studies in the whole-animal model system C. elegans, have shown that S. flexneri kills worms by an infection-like process that requires live bacteria, correlates with the bacterial accumulation in the intestine, and requires the S. flexneri virulence plasmid which is crucial for pathogenesis in mammalian systems; in addition, the elegans Src homology, non-receptor tyrosine kinase ABL-1, is required for S. flexneri-induced killing, as loss of ABL-1 through mutation, RNA interference(RNAi), or pharmacological inhibition via the inhibitor STI571, results in extension of nematode life span on S. flexneri." "Shigella flexneri" Paper_evidence "WBPaper00027729"
Different scenarios with Strain information:
- authors may not always mention the strain for the pathogen
- may reference an older paper and things could get messy from there, if I have to identify which exact strain they used based on the older paper.
In these cases, I'd like to use ?Species (when strain information is hard to find). Not sure it is absolutely important to curate to the level of strain! However when strain information is present, the following model lets me curate to that level of granularity:
?Gene Infection_model Required_for_infection_by_strain ?Strain XREF Host_gene_required_for_infection #Evidence Resistance_to_infection_by_strain ?Strain XREF Host_gene_required_for_resistance #Evidence Required_for_infection_by_species ?Species XREF Host_gene_required_for_infection #Evidence Resistance_to_infection_by_species ?Species XREF Host_gene_required_for_resistance #Evidence Infection_model_description ?Text ?Species #Evidence ?Strain Infection_model_info Host_gene_required_for_infection ?Gene XREF Required_for_infection_by_strain Host_gene_required_for_resistance ?Gene XREF Resistance_to_infection_by_strain Infection_model_description ?Text #Evidence ?Species Infection_model_info Host_gene_required_for_infection ?Gene XREF Required_for_infection_by_species Host_gene_required_for_resistance ?Gene XREF Resistance_to_infection_by_species Infection_model_description ?Text #Evidence
After discussions with the Caltech group, here is a more simplified model that takes into account that the same gene might be an infection_model for more than one type of pathogenic strain:
?Gene Infection_model Required_for_infection ?Text UNIQUE ?Strain XREF Host_gene_required_for_infection #Evidence Required_for_resistance ?Text UNIQUE ?Strain XREF Host_gene_required_for_resistance #Evidence ?Strain Infection_model_info Host_gene_required_for_infection ?Gene XREF Required_for_infection #Evidence Host_gene_required_for_resistance ?Gene XREF Required_for_resistance #Evidence Back to Disease and Drugs