muscle attachment site

[mellybelly]

We probably need a class for this, as currently we only record muscles being attached to skeletal elements, but not specifically where. Since muscle attachment sites are a) used extensively in character state descriptions, and b) there are many standard ones that have names, we should have a way to express this.

Some examples: sagittal crest deltopectoral crest trochanter on femur

[cmungall]

Can we auto-infer?

A naive logical def:

• 'muscle attachment site' EquivalentTo 'anatomical structure' and inverseOf(attaches_to) some 'muscle organ'

However, this would classify entire bones as being attachment site. Our current axioms say things like "muscle X" has_origin some "bone Y", where has_muscle_origin is a subproperty of attaches_to.

Presumably the attachment site is constrained to be the area where the bone attaches. Is this a flat site (sensu bfo) on the bone? Is the trochanter really an attachment site, or is the attachment site on the surface of the femur?

[mellybelly]

Yep exactly, we can auto-infer that there IS an attachment site, but not exactly where on the skeletal element.

There are multiple attachment sites on the trochanter. I think attachment sites (for tendons as well, I presume) can be convex, flat or concave?

We might be able to get around having "attachment sites" if we can add enough of the skeletal element parts and make the attach relations to those.

[cmungall]

A few other questions - this is a metazoan anatomy ontology; muscle applies across metazoans. Should we have generic attachment site and "muscle attachment site of bone".

Might be good to have FEED feedback here too.

Also - how does this related to muscle scar?

[cmungall]

[Term]
id: FBbt:00005089
name: muscle attachment site
namespace: fly_anatomy.ontology
def: "The place where the muscle system attaches to the body wall." [FBC:gg]
subset: FB_gloss  ! terms for FlyBase glossary page
subset: cur  ! cur
subset: hidden_assertion  ! Stanza contains hidden assertion.
is_a: FBbt:00100290  ! endocuticular specialization
relationship: connected_to FBbt:00005073 ! somatic muscle
relationship: part_of FBbt:00005069 ! muscle system

[mellybelly]

If we decide to implement muscle attachment sites, they should indeed be generic. Also, muscles in vertebrates can attach to cartilage and then there are tendon attachment sites on bones.

Muscle scars are basically attachment sites for muscles as seen in skeletal remains, though there may be other more general definitions. So 'muscle scar' would be a related synonym?

[cmungall]

From Robert D: "in fact, muscle never really attaches directly to bone. There is always a tendon and often the tendon attaches to periosteum and the periosteum attaches to the bone"

i.e.

has_part attaches_to part_of we are currently using attaches_to in a loose way (muscle to bone). We could have a formalism that uses a directly_attaches_to, which would hold between for example (not exclusively) tends and periosteum. We could then define the looser relation in terms of the chain above. Of course, this is just one kind of attachment