Semantic resources project/MouseModels/Bottom-up

= Sample Mouse Model =

Sample AD-related strain: B6.129-Psen1tm1Shn/J
 * Jax data sheet
 * MGI:87705 = J:40365
 * associated allele of interest = Psen1tm1Shn

Class: mouse (members are individual mice) Subclasses by restriction on phenotype, genotype, diet, supplier, lineage, etc. Class: mouse population ? the subject of statistical or predictive statements ("50% show this phenotype at 3 months" is not a property of individual, is it?) ''[Why wouldn't that just a be a "disposition" that inheres in each individual? Dispositions don't need to be deterministic, do they? -TWD]''

Cryopreserved: "Homozygous mutants die shortly after natural birth. Heterozygous mutants are viable and fertile." - so what are the mice that are preserved? presumably heterozygous? what does that say about the members of the "strain". They have a disposition to have offspring that ...

Provenance
Generation: N6+4pN1 (see the JAX Husbandry Page on Generation definitions)
 * "N6" -- 'N' is the number of "backcross generations," so this is six backcross generations.
 * "+" -- The location of the N6 before the '+' symbol indicates that these backcrosses occurred in the lab of the donating investigator.
 * "4" -- four additional (backcross?) generations took place after transfer to JAX.
 * "p" -- at this point, the mice were cryopreserved.
 * "N1" -- one additional backcross generation, at JAX (after cryo-preservation? Is this the generation they give you upon defrosting the mice?)

Donating Investigator: "Jie Shen, Harvard Med Sch/ Brigham Women's Hosp"
 * Jie Shen's faculty webpage
 * presumably, this is what the "Shn" fragment of the allele identifier is indicating: "Shen"
 * And indeed, a search of the ILAR Lab Codes indicates that this is the case -- although ILAR doesn't provide any re-usable URLs to link to, for specific labs.

This strain was originally reported in Shen et al. "Skeletal and CNS defects in Presenilin-1-deficient mice." Cell (May 1997)
 * PMID: 9160754
 * MGI:87705
 * There's also this ID: "J:40365", which is the "MGI Ref ID"

ES Cell Line: AB2.1
 * I can't find (yet) a good reference for this cell line.

Strain of Origin: 129S7/SvEvBrd-Hprt
 * This appears to be a strain that's not in the JAX system. See this bulletin on the 129 strain nomenclature, and the table at the end.
 * Here is a related strain, 129S7/SvEvBrd-Mt1tm1Bri Mt2tm1Bri/J, that is in the JAX system.
 * "Hprt" appears to be a gene name.

Genetic Content
The allele, "Psen1tm1Shn", stands for "first targeted mutation" of the Psen1 gene by the Shen lab.
 * Psen1 gene: MGI, OMIM, Entrez Gene (Human), Entrez Gene (Mouse)
 * Psen1 protein: Uniprot (Mouse), Uniprot (Human), Pfam, PRO (and there are at least six isoforms, which are not in PRO yet).
 * Psen1 appears to interact with the Notch signaling pathway.
 * The "tm1Shn" allele is a "neomycin cassette" that targets a key exon of Psen1.
 * A high-level description of this kind of knockout construct.
 * From JAX, the "allele type" is "Targeted (knock-out)".
 * "Exon 3 of the Psen1 gene, encoding the translation initiation codon, was deleted and replaced with a neomycin cassette. "
 * Northern blots were then used to check the Psen1 mRNA, westerns to check Psen1 protein, and (presumably) neomycin resistance was used to select the cells in the first place which had integrated the construct.

Phenotype
There's an implicit hierarchy of phenotypes asserted. For example, here is one sub-section:

skeleton phenotype abnormal axial skeleton morphology * axial skeleton is severely malformed  (MGI Ref ID J:40365) * caudal to the pelvis, the axial skeletal structure is completely missing  (MGI Ref ID J:40365) abnormal occipital bone morphology * occipital bones are underdeveloped  (MGI Ref ID J:40365)

So, "abnormal occipital bone morphology" phenotype is a kind of (subclass of?) the "skeleton phenotype". (Furthermore, a search reveals that terms like "abnormal occipital bone morphology" are reused across many mouse strains.)

Some searching in NCBO suggest that these may be controlled terms from the Mammalian Phenotype ontology. ''[Also, the fact that the phenotype list is labeled with "Mammalian Phenotype Terms" probably should have tipped me off. -TWD]''
 * MP on OBOFoundry
 * MGI has their own search form for MP, so yes: this is probably what they're using.
 * MPO:0005269 is the MPO term for "abnormal occipital bone morphology"
 * "any structural anomaly of the bone at the lower, posterior part of the skull"
 * is_a "abnormal neurocranium morphology", so that suggests that the JAX page is not displaying the full class tree.

The '*' starred lines are not general terms, but are summaries of findings taken from the reference given by the MGI Ref ID at the end of the line. So, for example, the phrase "occipital bones are underdeveloped" is taken from the longer passage,

The sternum of mutant mice is shorter and thicker relative to wild-type mice, and lacks intersternebral cartilage, by which ribs attach to the sternum. An  additional skeletal abnormality present in the mutant is underdevelopment of   the occipital bones. The bones of the limbs, pelvis, scapulae, and clavicles are morphologically normal in the PS1 mutant mice.

(pg. 631 of Shen et al., bottom right column)

So someone has read this paper, and manually curated this claim into the phrase, "occipital bones are underdeveloped." I (TWD) think this is probably ready for a SWAN-like representation, along with some proper provenance, right?
 * We could talk about these as dispositions: every mouse which is homozygous for this allele has a part (occipital bone) which bears a quality ("underdeveloped").
 * This restriction could be given an annotation as a means of sourcing it to a SWAN claim, where the claim makes clear that the statement "occipital bones are underdeveloped" is derived from the Shen et al. paper, but was made ("created" in SWAN terminology) by an MGI/JAX editor. Probably should check with someone (TM?) about who, exactly, is creating these statements.

This brings up a second distinction: some of the phenotypes that are present on this description apply to the actual mice that are part of the offer, and some descriptions apply to homozygous mice which could be (and in the past: have been) derived from these mice by mating.
 * How should this be represented?

Mammalian Phenotype Ontology
It's a collection of classes, arranged hierarchically using is_a relationships. There are no properties, no restrictions, no intersections. (There do appear to be some clases with multiple parent classes, eg.

[Term] id: MP:0000015 name: abnormal ear pigmentation def: "anomaly in the coloration of the ear" [MGI:cwg "Carroll-Ann W.    Goldsmith, Mouse Genome Informatics Curator"] is_a: MP:0002095 ! abnormal skin pigmentation is_a: MP:0002177 ! abnormal outer ear morphology

but most classes have a single parent.) It's not integrated with OBI -- one could imagine that most of these classes would be "qualities" in the BFO-sense, but at least some of the upper-level classes might be expressed as restrictions.

"abnormal outer ear morphology" subclassOf (inheres_in some (part_of some 'outer ear'))

or something along those lines.

I'm not sure what the plan is then, with respect to the MP. -TWD

= Sample Allele =

Sample AD-related allele report:

Psen1tm1Pcw

- can't find easily in Alzforum.

= Example =

JAR and TWD worked out an example, in diagrammatic form, of how we would like to represent a mouse strain, an allele containing a genetic modification, and an associated phenotype.