phi-base-use-case.ttl

@prefix res: <http://agrischemas.org/examples/dfw-hack-2019/resources/> .
@prefix agri: <http://agrischemas.org/> .
@prefix bioschema: <http://bioschemas.org/> .
@prefix schema: <http://schema.org/> .
@prefix obo: <http://purl.obolibrary.org/obo/> .
@prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> .
@prefix dc: <http://purl.org/dc/elements/1.1/> .


# This is the container that glue things together, based on PHIBase records.
#
# Host-pathogen interactions are a specific case of process, action, or alike
# We use schema:Action for that. This is somewhat close, not ideal, to what we mean, since 
# schema actions seem specific to human actions (eg, agent is restricted to Organization or Person).
# 
# Probably it requires some other kind of process/action, or a better definition of the existing Action.
# Bioschemas doesn't seem to have any mention of these concepts
#
# Clearly, this is relevant in the defition of pathways too (see the biomolecular use case).
#
# schema:Action can be redundantly added, to ease consumers like Google
res:phiInt001 a schema:Action, agri:HostPathogenInteraction;  
  # Similarly, this is redundant but could help
  schema:additionalType
    # host-pathogen interaction
    <http://purl.obolibrary.org/obo/OHPI_0000001>,
    # biological process involved in symbiotic interaction
    <http://purl.obolibrary.org/obo/GO_0044403>.

  # in schema.org, schema:object is synonym of an Action participant.
  # So, in this interaction a host, a pathogen and a pathogen's gene are the participants.
  # 
  # An alternative could be linking the gene to the pathogen and then link the gene only here
  # to the interaction.
  #
  schema:object 
    phiInt001_host,
    res:phiInt001_pathogen,
    res:gene_FGRAMPH1_01T00873;

  # Means we get these results from this interaction and the above participants
  #
  schema:result 
    # The resulting PHI Host-related Phenotype
    res:phiInt001_host_phenotype,
    # The resulting PHI pathogen-related Phenotype
    res:phiInt001_pathogen_phenotype;
  
  agri:evidence res:pmed_18943005;

  agri:evidence [
    a schema:PropertyValue;
    schema:propertyID "phiBase:experimental technique";
    schema:value "Gene disruption"
  ]
.


# A PHI host phenotype, resulting from some PHI interaction
#
res:phiInt001_host_phenotype a bioschema:Phenotype;
  # TODO: schema:role has been removed, so we propose to introduce this agri:role property
  # An alternative could be: additionalType: phiBase:HostPhenotype, but this.
  # 
  # A possible alternative could be <this> schema:roleName "host", which would maket the host
  # a role too (a bit dirty)
  # 
  agri:role [
    a schema:Role;
    schema:roleName "PHI Host Phenotype";
  ];
  
  schema:additionalProperty [
    a schema:PropertyValue;
    schema:propertyID "phiBase:Phenotype";
    schema:value "reduced virulence"
  ];

  # This could be as simple as a plain string, here we show how to attach annotations
  # such as ontology terms.
  #
  bioschema:associatedDisease [
    # As usually, we imagine this to be created from imported strings + annotations
    # like text mining annotation. Due to that, this is a property value, since we aren't
    # sure it's a DefinedTerm. As you can see, the latter can be added.
    #
    a schema:PropertyValue;
    
    dc:type obo:TO_0000663;

    # As you can see from the gene annotation below, 
    # these could be taken from resolving the dc:type, but they're often added to ease data
    # consumers
    schema:propertyID "disease name";
    schema:value "Fusarium Ear Blight";
  ];

  # Redundant (it's already in the Action above), but complies with bioschemas and could 
  # be useful
  bioschema:bioChemAssociation res:phiInt001_host;
.


# A PHI pathogen phenotype, resulting from some PHI interaction
#
res:phiInt001_pathogen_phenotype a bioschema:Phenotype;
  # TODO: see notes above
  # 
  agri:role [
    a schema:Role;
    schema:roleName "PHI Pathogen Phenotype";
  ];
  
  schema:additionalProperty [
    a schema:PropertyValue;
    schema:propertyID "phiBase:Postpenetration defect";
    schema:value "yes"
  ]

  # As above, redundant since it's already in the Action above.
  bioschema:bioChemAssociation res:phiInt001_pathogen;

  # You might want this redundancy too, to quickly link a gene to a phenotype
  bioschema:bioChemAssociation res:gene_FGRAMPH1_01T00873;
.


# The host details
#
res:phiInt001_host a schema:BioChemEntity;
  schema:name "Triticum aestivum";
  dc:type <http://purl.bioontology.org/ontology/NCBITAXON/4565>;

  # See notes above
  agri:role [
    a schema:Role;
    schema:roleName "host";
  ]
.

# The pathogen details
#
res:phiInt001_pathogen a schema:BioChemEntity;
  schema:name "Fusarium graminearum";
  # As per bioschemas
  bioschema:taxonomicRange <http://purl.bioontology.org/ontology/NCBITAXON/5518>;
  # Possibly, an additional annotation that is more standard
  dc:type <http://purl.bioontology.org/ontology/NCBITAXON/5518>;

  # See notes above
  schema:role [
    a schema:Role;
    schema:roleName "pathogen";
  ]
.


# The gene and protein details
#
res:gene_FGRAMPH1_01T00873 a bioschema:Gene;
  schema:identifier "FGRAMPH1_01T00873";
  schema:name "Transducin beta-subunit", "TBL1";
  bioschema:encodes res:protein_A0A1C3YHH5;
  dc:type obo:GO_0009405, obo:GO_0044412 .

res:protein_A0A1C3YHH5 a bioschema:Protein;
  dc:source "UniProt"
.

# the publication evidence
#
res:pmed_18943005 a agri:ScholarlyPublication;
  dc:title  "Random Insertional Mutagenesis Identifies Genes Associated with Virulence in the Wheat Scab Fungus Fusarium graminearum.";

  schema:abstract "Fusarium graminearum is an important pathogen of small grains and maize in many areas of the world. To better understand the molecular mechanisms of...";

  agri:pmedId "18943005" ;

  # This should be an ISO 8601 string. This standards accepts dates with year only, however it
  # isn't clear if they should be interpreted as 'day/month is unknown', or 'the year is referred by generically', or
  # 'Jan 1st is implicit'
  schema:datePublished  "2005-07"
.