taxis

[gocentral]

Looking at the definitions, I think 'taxis' and all its children should be moved under "behavior"

-Taxis (GO:42330): The directed movement of a cell or organism in response to an external stimulus.

Behavior (GO: 7610)The specific actions or reactions of an organism in response to external or internal stimuli. Patterned activity of a whole organism in a manner dependent upon some combination of that organism's internal state and external conditions.

---

Moreover, I would suggest merging -phototaxis and phototactic behavior -gravitaxis and gravitactic behavior

Pascale

Reported by: pgaudet

Original Ticket: "geneontology/ontology-requests/2890":https://sourceforge.net/p/geneontology/ontology-requests/2890

[gocentral]

Logged In: YES user_id=631592

How does taxis differ from locomotory behavior? I actually think taxis includes part of a cell. For example, I think the directed movement of an axonal growth cone is a taxis, but it is not a behavior. So, I actually think the definition of taxis is too restrictive.

Now we will really see if Doug is listening!

David

Original comment by: ukemi

[gocentral]

Logged In: YES user_id=835115

I believe taxis is directed movement [of a whole organism?] toward or away from a stimulus...as we have it defined in the GO. phototaxis is movement toward (or away from) light...etc... The current definition of phototaxis could be modified to clarify this explicitly.

It seems like *taxis terms should be children of 'locomotory behavior', only if they refer to movement of the whole organism.

I think if a cell bent or extended a process towards a chemical stimulus for example, that would be chemotropism rather than chemotaxis, wouldn't it?.

Original comment by: doughowe

[gocentral]

Logged In: YES user_id=631592

This makes sense to me. So the taxis terms should somehow be related to locomotion.

David

Original comment by: ukemi

[gocentral]

Logged In: YES user_id=631592

I meant to say locomotory behavior last night. Is there any way we can role these items into one? I think we should tackle the locomotion, locomotory behavior and taxis issues in one sourceforge item.

I'm wondering if you guys think it would be worth it for me to expand on the locomotion part of the graph. We have walking behavior, but would it be worthwhile to add things like swimming, sinusoidal movement, ciliary/flagellar movement and such?

David

Original comment by: ukemi

[gocentral]

Logged In: YES user_id=803460

Hi David,

I think it would help to add other behaviors.

A couple of things: -intuitively, I find it difficult to distinguish between "xx ' and xx'behavior terms and that's why I would like to merge them (including locommotion and locomotory behavior).

-about behavior: I remember some sort of disucssion about certain kinds of locomotions not being 'in response to a stimulus'. For example walking and swimming--- this stimulus is different from tactic behaviors. Would it be possible to change the definition of 'behavior' from 'in response to a external or internal stimuli" to something that rather reflects the fact the the action is volountary?

-regardless of the above, it seems like most children of 'locomotion' really are behaviors (ie, that they are occuring in response to a stimulus).

-As far as I can see, the only reason to distinguish between 'locomotion' and 'locomotory behavior' is 'cell motility'. If you remove 'stimulus from the 'behavior' definition, then that term can also be in the same node. Otherwise, it could stay on its own as a cellular physiological process.

Pascale

Original comment by: pgaudet

[gocentral]

Logged In: YES user_id=631592

Hi Pascale,

An example of the difference between sinusoidal locomotion and sinusoidal locomotory behavior. A worm normally cruises around on a petri dish looking for food. If you tap the worm on the nose with a small stick, it backs up. There are mutants that can cruise around on the plate just fine, but when you tap them on the nose, they don't back up. Then, there are other worms that are uncoordinated that simply don't move well. Would we like to distinguish between these? Maybe we should get a worm person in on this too.

I've written to Kimberly.

David

Original comment by: ukemi

[gocentral]

Logged In: YES user_id=631592

Hi Everyone,

So now I am thinking that maybe locomotory behavior and locomotion should be separate. I don't think we should change the definition of behavior. But, I think that locomotory behavior should be:

The specific actions or reactions of an organism in response to external or internal stimuli that result in a change in locomoter activity. These could include the cessation, initiation or change in direction of locomotion.

I think what I am getting at is that when I am walking I am walkingthat doesn't change. When I see a car coming, I stop walking. That is the behavior. Does this make sense?

David

Original comment by: ukemi

[gocentral]

Logged In: YES user_id=886166

I'm the guy who originally got these two terms made distinct, so I should probably explain myself...

The original reason for these two terms being separated is that it's genetically possible to separate two different capacities: [1] the ability to move at all whatosoever (which is lost in worm, fly, and human mutations causing total paralysis), versus [2] the ability to not merely eke out some sort of motion but to do it in the fully healthy way of a well-adapted wildtype organism (which is lost in worm, fly, and human mutations causing uncoordinated movement).

So both "locomotory behavior" and "locomotion" are considered to be defective in completely paralyzed (Prz) worms (e.g., from RNAi), but only "locomotory behavior" is considered defective in motile-but-uncoordinated (Unc) worms (from RNAi or mutations).

This phenotypic distinction can be seen in mutations from three rather different organisms (worm, fly, and human); it doesn't depend on the peculiarities of any one animal phylum. So it seemed, and still seems, reasonable to me infer genetically dissectable differences in biological processes from the mutant phenotypes.

Original comment by: emsch

[gocentral]

Logged In: YES user_id=886166

About not all motion being "in response to a[n external] stimulus": that has been known by ethologists for some time. (E.g., it was already standard knowledge in first-year-grad-school ethology class at Caltech by the mid-1980s.) That's why, when I wrote the definition of behavior for GO, I was careful to make it clear that not all behavior was driven by external stimuli:

"The specific actions or reactions of an organism in response to external or internal stimuli. Patterned activity of a whole organism in a manner dependent upon some combination of that organism's internal state and external conditions."

Original comment by: emsch

[gocentral]

Logged In: YES user_id=803460

Hello again,

I apologize if that entry is starting to cover too many different things-- but I do have one more comment to make.

First, thanks to Erich for explaining the distinction between 'locomotion' and locomotory behavior'.

It still seems counter-intuitive to describe these two processes as 'lomomotion' and 'locomotory behavior'. Do papers dealing with these mutants use these terms like this?

It seems like the disticntion is between the 'detection' and the 'response'. There was a SF item recently (1290786), following which Amelia (reluctantly) added this to the GO documentation: http://www.geneontology.org/GO.process.guidelines.shtml?all\#stim

To me it would be more logical to have : --locomotion ---coordination of locomotion [GO:new?] ----forward locomotion ----backward locomotion ------backward locomototion in response to tapping on the nose [GO:new?]

--response to external stimulus ----response to tapping on the nose of the worm ------backwards locomototion in response to tapping on the nose

This kind of structure seems (maybe just to me) more consistent with other response-type terms.

That's for locomotion- if we dont agree we ca leave that as such for now, and just fix the 'taxis' terms.

Pascale

Original comment by: pgaudet

[gocentral]

Logged In: YES user_id=893454

Below is how I feel about the locomotion and locomotory behavior terms:

I would agree with Erich that there is a difference between the ability to locomote at all, and the ability to locomote in a coordinated way.

However, I would switch the parent-child relationship of locomotion and locomotory behavior. In GO right now (in brief):

locomotory behavior -- locomotion -- regulation of locomotion

I put in the below two terms, and it was modeled according to the way terms are usually input in the GO. As an example, look at cell migration and regulation of cell migration terms, and their associated definitions.
However, when I input a definition for regulation of locomotion, I should have used the following definition: Any process that modulates the frequency, rate, extent, or coordinated patterns of activity of any organism.

I consider a protein that renders a worm paralyzed when it is knocked out to be required absolutely for locomotion, and that includes both coordinated and uncoordinated locomotion. I would merge regulation of locomotion with locomotory behavior. And model the relationships in this way:

-- locomotion -- regulation of locomotion

With respect to the actual processes we are talking about I think this type parent-child of relationship makes the most sense. There are many worm proteins, which when knocked out, render a worm paralyzed. However, less severe alleles (partially functional proteins) render a worm capable of some kind of movement but not in a coordinated fashion (i.e., egl-30, unc-13).

Carol Bastiani

Original comment by: carolbas

[gocentral]

Logged In: YES user_id=436423

Has there been a decision yet about what to do for locomotion and locomotory behavior? As I said in SF 1177217, the GO Eds will implement whichever version works best for those who annotate organisms that move, but I don't want to make a decision for which bench work on yeast is totally irrelevant :)

m

Original comment by: mah11

[gocentral]

Logged In: YES user_id=631592

Midori,

I certainly think that taxis should be a locomotory behavior. I also think that the merges suggested should be implemented. As for locomotion being a child of locomotory behavior. I think that from the thread below: 1) it is clear that not all locomotion is a part of locomotory behavior, 2)it is not clear that the elements of locomotion that are a part of locomotory behavior are any different from just being able to locomote in any circumstance. I still think it would be best to just separate locomotion from locomotory behavior. Maybe folks could just vote on the two options.

David

Original comment by: ukemi

[gocentral]

Logged In: YES user_id=803460

Hello,

I am happy to do what David says. I think he is supporting Carol's proposal to make locomotion a parent of locomotory behavior. Also, I would support the merge of 'locomotory behavior' with 'regulation of locomotion', I think that in essence these two terms descibe the same thing.

Thanks, Pascale I would prefer that to making locomotion completely separate.

Original comment by: pgaudet

[gocentral]

Logged In: YES user_id=886166

It seems to me that the genetic data Carol is citing actually support keeping the basic current relationship, in which locomotion is a child of locomotory behavior, but reclassifying this parent-child relationship from [is-a] to [part-of]. In the null alleles, the process of locomotion is reduced to zero, making other locomotory behaviors than mere locomotion impossible; in the hypomorphic alleles, basic locomotion is left intact but other subprocesses of locomotory behavior are now impaired. Locomotion is necessary, but not sufficient, for completely normal locomotory behavior -- which is what motivated the current parent->child relationship, but which may not be as well captured by a parent-[is-a]->child relationship as it would be by a parent-[part-of]->child relationship.

Original comment by: emsch

[gocentral]

Logged In: YES user_id=803460

Hello again,

I dont really care either way. But if we leave the graph as it is, we need to fix some definitions.

Currently: -locomotion: Performing basic movements of any kind. -locomotory behavior: The abilities associated with normal movement, including both general motion and specifically coordinated patterns of activity.

Which (as far as I understand) makes 'locomotory behavior' is more specific than 'locomotion', and is inconsistent with their position in the graph.

Pascale

Original comment by: pgaudet

[gocentral]

Logged In: YES user_id=893454

I think that Pascale and I are basically thinking about this in the same way. As defined, locomotion should be a parent of locomotory behavior.

Even if locomotory behavior and regulation of locomotion are not merged, as I had suggested, locomotion (performing basic movements of any kind) should be a parent of any terms that capture coordination of movement, direction of movement, speed of movement, etc.

Original comment by: carolbas

[gocentral]

Logged In: YES user_id=886166

The definitions may indeed need rewriting if people genuinely find them confusing. Moreover, I think one can make a good argument for changing the nature of the parent-child relationship; I don't remember any more why I chose an is-a parent-child between "locomotory behavior" and "locomotion", but in retrospect it seems perhaps mistaken.

However, I can definitely say that, as originally intended, "locomotion" was meant to represent core competence to move at all (lost in fully paralytic mutants or RNAi animals), whereas "locomotory behavior" was meant to represent a broader set of abilities including not only the ability to move but the ability to move in a fully normal manner, typical of the wild-type animal. So I find it easy to see how the relationship could be changed from:

"locomotory behavior" --parent-[is-a]-child--> "locomotion"

to

"locomotory behavior" --parent-[part-of]-child--> "locomotion"

But I have to say, I find it very difficult to see how the relationship could sensibly be reversed to:

"locomotion" --parent-[is-a]-child--> "locomotory behavior"

given that this would make the ability to make completely normal, complex movements an instance of the ability to move at all. It seems to me that normal organismal movements require considerably more biological processes (abilities) than the bare ability to eke out some sort of spatial translocation alone.

Original comment by: emsch

[gocentral]

Logged In: YES user_id=436423

I certainly see Erik's point -- locomotory behavior does include locomotion, so it makes sense for a locomotion term to be part_of locomotory behavior (not is_a). Whether we can simply have the parentage

- locomotory behavior -- [p] locomotion

depends on whether it will violate the True Path Rule or not. The above arrangement implies that all locomotion (i.e. every instance of an organism moving under its own power) is also locomotory behavior. If so, we're fine. But if there is any movement that is locomotion but not locomotory behavior, we'll have to have something like this:

- locomotion -- [i] locomotion during locomotory behavior [or 'behavioral locomotion' or 'locomotion component of locomotory behavior' or some other awkward wording]

- locomotory behavior -- [p] locomotion during locomotory behavior [yada yada]

David made a comment in this thread a few days ago to the effect that not all locomotion qualifies as behavior, which would imply that we'd need the second, more complicated, structure if we want to maintain a connection in the ontology between locomotion and locomotory behavior.

Could those whose favorite organisms actually move please vote (or arm- wrestle, or whatever it takes to decide) and let me know what to implement?

Thanks! m

Original comment by: mah11

[gocentral]

Logged In: YES user_id=893454

My vote is: --locomotion -- locomotory behavior

An organism cannot exhibit any kind of locomotory behavior if it cannot locomote at all. A paralyzed organism does not exhibit locomotory behaviors, and so the concept of locomotion naturally encompasses evey kind of locomotory behavior, as well as coordination, direction of movement, etc. Locomotion is the more general term.

When a protein is annotated to the process of locomotion, I would assume that that protein is neccessary for locomotion of any kind. If instead we say a protein only affects locomotory behavior, then I would assume that its absence does not render an organism paralyzed, but instead affects specific aspects of locomotion, like coordination or locomotory behavior (however we choose to define it).

Specific behaviors that happen during locomotion are processes more specific to the general concept of locomotion.

Carol

Original comment by: carolbas

[gocentral]

Logged In: YES user_id=886166

Well, I've already voted. I think we need to get other people besides Carol and me to read this thread and collectively decide which viewpoint makes the most sense to them.

Original comment by: emsch

[gocentral]

Logged In: YES user_id=803460

Erich,

Can you explain and maybe give a couple of specific examples for "locomotory behavior"? What is 'broader than locomotion', exactly? You write, 'the ability to move in a fully normal manner'. How is that different from 'regulation of locomotion'?

Of course, Dictyostelium has much simpler locomotory behavior so I am probably just missing the point. To give you an example, with respect to chemotaxis of cells towards a chemoattractant: we have mutants that are unable to move at all, these we annotate to "cell motility" (which I consider the cellular equivalent of "locomotion"). Then other mutants can move, but do not respond to a chamoattractant properly. These get annotated to "chemotaxis", which is "The directed movement of a motile cell or organism in response to a specific chemical concentration gradient. (...)".

Of course the normal animal can move and also it can move in response to the chemoattractant. But having the ability to respond to a chemoattractant is only part of the ability to move, as some mutants can respond only to certain chemoattractants, or can respond to other stimili, such as light. To me, though, this is not 'the broader ability to move', but rather, the narrower ability to respond to specific stimuli. I thought Davis's "tapping on the nose" example was comparable.

Perhaps we should send this to the GO list for comments?

Pascale

Original comment by: pgaudet

[gocentral]

Logged In: YES user_id=886166

Normal, wild-type, healthy locomotory behavior of a worm consists of a graceful undulating slither over solid media, or a vigorous whipping swim motion in liquids. Until very recently, all studies of mutant movement in worms were of the movement over a solid substrate (culture plate) and not swimming. Movement over a solid surface is indeed regulated in response to stimuli (as Dictyostelium chemotaxis is), and indeed there are sensory mutants that block the adjustment of movement by chemical, mechanical, and other stimuli. However, what I had in mind with "locomotory behavior" is the entire, complex, preprogrammed set of detailed neural and muscular activities that comprises a "simple" clean worm movement. In the studies of ethology that have been done on other organisms these "simple" movements actually prove quite challenging and complicated to understand. For instance, even 20 years ago there were ethologists trying to understand how locomotion was programmed in things like stick insects, or swimming in fish (this led to interest in things like the Mauthner neuron).

What is broader than locomotion about locomotory behavior is that, for an organism to have optimal fitness in its native environment, it isn't sufficient for it to be able to move in some sort of way -- it actually needs to have some stereotypical motions so well tuned that it can do them at great length and with minimum wasted motion. Being able to move is necessary but not sufficient for being coordinated in these basic motions.

The instances of mutations in C. elegans that are perfectly able to move but are uncoordinated in ways that would kill them in nature are diverse. There's a mutant that is totally normal except for its ability to back up properly in response to a bump on its nose -- it clearly senses the bump (stops moving) but simply does not have the mojo to do a backward motion (even though it is thoroughly able to move forward). There's another two genes (unc-22 and lev-11) which, when mutated, result in chronic spasms -- the worms literally twitch all day. They can and do move, and can and do adjust their movement to different stimuli, but left to themselves they incessantly twitch like the worst coffee nerves ever. There are any number of uncoordinated mutants that have some sort of subtle defect in their otherwise healthy motion -- they can move, adjust movement, etc., but there's just some specific thing gimpy about them (weird kinks in the normal sinusoidal motion; coiling up; irregular waveforms or frequencies of sinusoidal motion; "jerky ratchet-like movement"; shrinking on itself, or constitutively hypercontracted; loose "loopy" sinusoids; "slightly rippling movement"; "shaker" spasms, more contracted than "twitcher" spasms ... you name it, there's some worm mutant that probably has it).

In short, it seems to me that the number of things that the genome has to get right in order for a worm to move properly is visibly more complex than the number of things it needs to get right to have a worm that isn't just paralytic. And this isn't simply a function of external regulation of the movement by outside stimuli, but of the genome building a very complicated set of gears and triggers just for the worm to be able to get fro one end of the plate to the other without slipping on an invisible banana peel.

That all being said, I agree that it'd be good to submit this question to the GO listserv.

Original comment by: emsch

[gocentral]

Logged In: YES user_id=436423

Let's see if I can restate the case for the arrangement with locomotory behavior as a child of locomotion (note: I'm not taking sides, I'm just trying to clarify some of the reasoning):

All of the features that distinguish locomotory behavior from any old locomotion can be thought of as additional criteria or conditions, making the meaning of the behavior term more specific. In other words, 'locomotory behavior' necessarily includes the ability to move, plus all the refinements such as responding to stimuli, coordinating movements, etc. In this scenario, just plain 'locomotion' refers to movement with OR without the additional features that make for locomotory behavior.

I've also dropped a line to the mailing list.

m

Original comment by: mah11

[gocentral]

Logged In: YES user_id=893454

Hi Midori, Thanks for looking at this from different pperspectives. Sorry to pipe in again; just wanted to clarify my perspective.

The biological process of wild-type locomotion for a wild-type animal in a normal environment involves locomotion in a coordinated fashion, forward and backward movement, response to stimuli, etc. If a protein is involved in the process of locomotion, then one should assume that protein is involved with all aspects of wild-type locomotion. Thus, in the absence of that protein, the organism would be rendered immobile.

A protein that is only involved with coordinated movement, or the duration of forward movement (i.e., 11545720), or response to stimuli is involved with one or more (certainly) very complicated aspects of wild-type locomotion that can be separated from just moving, and an annotation to any of these subterms would indicate that the organism can still move in the absence of this protein.

So, the process of wild-type locomotion for a wild-type animals would encompass the ability to move, plus all of the refinements. The refinements are subsets of locomotion and could be classified as locomotory behavior.

Thus, this does not creat any problems with true path rules (as you had outlined below if we model it oppositely) because whenever an organism exhibits locomotory behavior, it is locomoting. And this makes sense because in the absence of locomotion, there is no locomotory behavior.

Midori wrote: >depends on whether it will violate the True Path Rule or not. The above >arrangement implies that all locomotion (i.e. every instance of an >organism >moving under its own power) is also locomotory behavior. If so, we're >fine. >But if there is any movement that is locomotion but not locomotory >behavior, we'll have to have something like this

Original comment by: carolbas

[gocentral]

Logged In: YES user_id=436423

anyone else want to vote or comment?

Original comment by: mah11

[gocentral]

Logged In: YES user_id=803460

I think phototaxis/phototactic behavior and gravitaxis and gravitactic behavior can be merged.

Maybe we can wait for the next GO meeting to discuss locomotion, locomotory behavior and regulation of locomotion?

Pascale

Original comment by: pgaudet

[gocentral]

Logged In: YES user_id=436423

OK, I've done the -tactic behavior/-taxis merges. I would love to close this item, but will leave it open for a little while longer because it contains so much discussion that's relevant to locomotion generally. Please use SF 1177217 and SF 1307099 for all further comments!

m

links:

http://sourceforge.net/tracker/? group_id=36855&atid=440764&func=detail&aid=1177217

https://sourceforge.net/tracker/index.php? func=detail&aid=1307099&group_id=36855&atid=440764

Original comment by: mah11

[gocentral]

• assigned_to: nobody --> gomidori

Original comment by: mah11

[gocentral]

Logged In: YES user_id=436423

I think I've got the other things (SF 13070099, 1177217) fixed now. Please take a look and see if anything else needs to be done. If not, I'll close this (at last!).

Original comment by: mah11

[gocentral]

Logged In: YES user_id=436423

No further comments (or complaints) have come in, so I'm closing these.

m

Original comment by: mah11

[gocentral]

• milestone: --> TermGenie
• status: open --> closed-fixed

Original comment by: mah11