Tanya... can you shed any light on the plant structure of GO:0048196? It's defined as:
The matrix external to the cell, composed of the cell wall and middle lamella.
Thanks,
Becky
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Note that this term does not have 'extracellular region ; GO:0005576' as a parent because in plants the extracellular matrix is considered part of the cell.
However, right now, it does have 'extracellular region ' as a parent. I suggest solving this problem by obsoleting 'extracellular matrix' as this is an artificial grouping term as currently defined, grouping by string and not by location:
A structure lying external to one or more cells, which provides structural support for cells or tissues; may be completely external to the cell (as in animals) or be part of the cell (as in plants).
Instead, we'd just retain the two children terms:
'middle lamella-containing ECM' which would be part of the cell and 'proteinaceous ECM' which would be part of the extracellular region , no more mixing of external to cell and part of cell.
We'd have to disentangle the other ECM terms (binding, assembly, disassembly, regulation,...) to figure out which were plant and which were animal or just create specific ones for each but I think this would serve to clarify the issue, not confuse it.
Thoughts?
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I think grouping classes are very useful when there is some shared biological characteristic, be it functional, structural, homology, etc. I think it gets very dubious with terms like "epidermis development" - I would be in favour of splitting that one.
I always thought that ECM, whilst different in constituents in plants and animals, seemed like a reasonable grouping class, but given what you say I think a split might be in order. This would solve the part/not-part issue. The current def doesn't really make sense does it? The implication is that plant ECM is simultaneously outside and inside the cell. It should really say "external to the plasma membrane" (or something)
Seems like a pretty big change though. And what about dicty, fungi..?
Didn't we discuss this at one of the GO meetings? Michael had an analogy about waxing a car (or waxing something).
An alternative to splitting the class is to re-think "extracellular", and perhaps subdivide this - e.g. extrernal to plasma membrane vs external to the cell+its parts.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I was just looking at annotations to 'ECM' and 'proteinaceous ECM'. From that glance, it seems that most of the ECM annotations are for mouse/human proteins and could easily be transferred down to 'PCM'. There's one TAIR annotation and possibly a few Dicty ones.
>The current def doesn't really make sense does it? The implication is that
>plant ECM is simultaneously outside and inside the cell. It should really
>say "external to the plasma membrane" (or something)
Yes.
If we want a grouping term for the plant and animal ECMs, perhaps ECM needs to be a direct child of cellular component and the children terms point to the appropriate cell part vs. extracellular region part parents.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Let's remove that disjointness axiom for now. If you like, we can discuss as a case in point in Hinxton. What if this axiom existed? How would we identify the offending terms? Good practice, I think. Maybe we'll even be able to tease out the biology.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
all of these fall in the same category of error
GO:0005607-laminin-2 complex
GO:0009530-primary cell wall
GO:0009531-secondary cell wall
GO:0020005-symbiont-containing vacuole membrane
GO:0020006-symbiont-containing vacuolar membrane network
GO:0048196-middle lamella-containing extracellular matrix
GO:0048217-pectic matrix
GO:0048222-glycoprotein network
GO:0048223-hemicellulose network
GO:0048224-lignin network
GO:0048225-suberin network
The term causing most of the problems is:
middle lamella-containing extracellular matrix ; GO:0048196
is_a: extracellular matrix
is_a: external encapsulating structure.
Tanya... can you shed any light on the plant structure of GO:0048196? It's defined as:
The matrix external to the cell, composed of the cell wall and middle lamella.
Thanks,
Becky
Reviewing this, I see the comment:
Note that this term does not have 'extracellular region ; GO:0005576' as a parent because in plants the extracellular matrix is considered part of the cell.
However, right now, it does have 'extracellular region ' as a parent. I suggest solving this problem by obsoleting 'extracellular matrix' as this is an artificial grouping term as currently defined, grouping by string and not by location:
A structure lying external to one or more cells, which provides structural support for cells or tissues; may be completely external to the cell (as in animals) or be part of the cell (as in plants).
Instead, we'd just retain the two children terms:
'middle lamella-containing ECM' which would be part of the cell and 'proteinaceous ECM' which would be part of the extracellular region , no more mixing of external to cell and part of cell.
We'd have to disentangle the other ECM terms (binding, assembly, disassembly, regulation,...) to figure out which were plant and which were animal or just create specific ones for each but I think this would serve to clarify the issue, not confuse it.
Thoughts?
I think grouping classes are very useful when there is some shared biological characteristic, be it functional, structural, homology, etc. I think it gets very dubious with terms like "epidermis development" - I would be in favour of splitting that one.
I always thought that ECM, whilst different in constituents in plants and animals, seemed like a reasonable grouping class, but given what you say I think a split might be in order. This would solve the part/not-part issue. The current def doesn't really make sense does it? The implication is that plant ECM is simultaneously outside and inside the cell. It should really say "external to the plasma membrane" (or something)
Seems like a pretty big change though. And what about dicty, fungi..?
Didn't we discuss this at one of the GO meetings? Michael had an analogy about waxing a car (or waxing something).
An alternative to splitting the class is to re-think "extracellular", and perhaps subdivide this - e.g. extrernal to plasma membrane vs external to the cell+its parts.
I was just looking at annotations to 'ECM' and 'proteinaceous ECM'. From that glance, it seems that most of the ECM annotations are for mouse/human proteins and could easily be transferred down to 'PCM'. There's one TAIR annotation and possibly a few Dicty ones.
>The current def doesn't really make sense does it? The implication is that
>plant ECM is simultaneously outside and inside the cell. It should really
>say "external to the plasma membrane" (or something)
Yes.
If we want a grouping term for the plant and animal ECMs, perhaps ECM needs to be a direct child of cellular component and the children terms point to the appropriate cell part vs. extracellular region part parents.
any chance this can be done before the hinxton workshop? If not I'll remove the disjointness axiom.
Let's remove that disjointness axiom for now. If you like, we can discuss as a case in point in Hinxton. What if this axiom existed? How would we identify the offending terms? Good practice, I think. Maybe we'll even be able to tease out the biology.
I think this has now been addressed by the changes made to ECM. Please reopen if necessary.