NTRs: behavior and human activity

[cmungall]

We should add behavior and link to NBO. We should follow NBO and make this a neurologically influenced behavior and not behavior in the very general sense.

We also need to think about where human activities like walking, smoking, going to the gym, etc fit in. I don't think they are in scope for NBO as these are not things a neurobiologist studies, and I don't think we should even put these in OBO as they are not biological. But NCIT has a decent starting vocabulary - see https://github.com/obo-behavior/behavior-ontology/issues/109

behavior is also mentioned in #42

[hoganwr]

I've been discussing a similar issue with Chris Stoeckert on the Apollo-SV listserv: https://github.com/ApolloDev/apollo-sv/issues/199

We're happy to model human activity in OMRSE, perhaps as some subtype of planned process.

On Tue, Mar 23, 2021 at 4:01 PM Chris Mungall @.***> wrote:

We should add behavior and link to NBO. We should follow NBO and make this a neurologically influenced behavior and not behavior in the very general sense.

We also need to think about where human activities like walking, smoking, going to the gym, etc fit in. I don't think they are in scope for NBO as these are not things a neurobiologist studies, and I don't think we should even put these in OBO as they are not biological. But NCIT has a decent starting vocabulary - see obo-behavior/behavior-ontology#109 https://github.com/obo-behavior/behavior-ontology/issues/109

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/OBOFoundry/COB/issues/156, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAJR55XTEK3IUJTVLIAKS5DTFDXTDANCNFSM4ZV3722Q .

[dosumis]

We should add behavior and link to NBO. We should follow NBO and make this a neurologically influenced behavior and not behavior in the very general sense.

👍

We also need to think about where human activities like walking, smoking, going to the gym, etc fit in. I don't think they are in scope for NBO as these are not things a neurobiologist studies, and I don't think we should even put these in OBO as they are not biological.

Lots of need for these for GWAS and whole genome sequencing association data. They need a reliable home. Some potential issues with boundaries here. "Going for a walk" may be a planned process, but walking itself is a behaviour controlled by particular circuits. And how do I record the behavioral results of messing with decision making circuits in the brain of an animal (lots of this type of working happening in Drosophila right now). Do we make a distinction between human planning and animal decision making?

[cmungall]

Just upping this as it came up in @hoganwr's ICBO talk. I think people are using GO hevavior to categorize processes that are not genetically determined.

Defining behavior is hard. See: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2760923/

I suggest the following as a pragmatic approach:

• COB: process
  • COB(new): behavioral action or activity of an organism
    • going for a walk
    • taking a class
    • aggressive actions or activities
      • COB/GO: biological process: genetically programmed
    • GO(+COB-new?):behavior genetically programmed
      • GO:learning or memory
      • GO:locomotory behavior
      • GO:aggressive behavior

Here the two classes would be class-disjoint. They could be linked by parthood relation at the domain ontology level. E.g with @dosumis' example, the NCIT:going-for-a-walk activity may have a has-part to GO:walking/locomotory behavior. COB would be neutral about when this would be done.

This approach can be criticised in that it can be really hard to know the boundaries. There will always be debates as to the extent of genetic influence on behavior, and things vary by species. We may end up with a bit of concept duplication and artificial labels to distinguish terms in the two hierarchies.

But it's very pragmatic. If there is evidence for the involvement in genes then the concept can be added to GO, if there are annotations. Otherwise keep it outside.

There is definitely much to be improved. I think the labels for GO behaviors should probably be modified to make their restricted sense clear. But my proposal here can go ahead without being blocked on this.

I don't know what that means for NBO, but it's inactive, and not really a COB problem. We just need a solution for people who need "behaviors" that don't belong in GO.

[hoganwr]

+1

I support this approach.

Matt, Clint, and I have done some work defining “human activity” but it has already been criticized on at least one issue thread (for some ontology) as being too species specific.

I think it needs to be more general as you say. This would be very helpful in a number of areas for OMRSE and other ontologies we’re working on.

On Tue, Sep 27, 2022 at 2:59 PM Chris Mungall @.***> wrote:

Just upping this as it came up in @hoganwr https://github.com/hoganwr's ICBO talk. I think people are using GO hevavior to categorize processes that are not genetically determined.

Defining behavior is hard. See: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2760923/

I suggest the following as a pragmatic approach:

• COB: process
  • COB(new): behavioral action or activity of an organism
    • going for a walk
    • taking a class
    • aggressive actions or activities
  • COB/GO: biological process: genetically programmed
    • GO(+COB-new?):behavior genetically programmed
      • GO:learning or memory
      • GO:locomotory behavior
      • GO:aggressive behavior

Here the two classes would be class-disjoint. They could be linked by parthood relation at the domain ontology level. E.g with @dosumis https://github.com/dosumis' example, the NCIT:going-for-a-walk activity may have a has-part to GO:walking/locomotory behavior. COB would be neutral about when this would be done.

This approach can be criticised in that it can be really hard to know the boundaries. There will always be debates as to the extent of genetic influence on behavior, and things vary by species. We may end up with a bit of concept duplication and artificial labels to distinguish terms in the two hierarchies.

But it's very pragmatic. If there is evidence for the involvement in genes then the concept can be added to GO, if there are annotations. Otherwise keep it outside.

There is definitely much to be improved. I think the labels for GO behaviors should probably be modified to make their restricted sense clear. But my proposal here can go ahead without being blocked on this.

I don't know what that means for NBO, but it's inactive, and not really a COB problem. We just need a solution for people who need "behaviors" that don't belong in GO.

— Reply to this email directly, view it on GitHub https://github.com/OBOFoundry/COB/issues/156#issuecomment-1259925815, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAJR55UCF3KGCS2OTBXMUO3WAM7Z5ANCNFSM4ZV3722Q . You are receiving this because you were mentioned.Message ID: @.***>

[bpeters42]

I am all in favor of giving these processes a home in COB, but would like to see the connection to 'planned process' being worked out well. It seems that with organism behavior, there typically is an 'objective', even if there isn't a full scale plan, so that needs to be made compatible.

• Bjoern

On Tue, Sep 27, 2022 at 12:22 PM Bill Hogan @.***> wrote:

+1

I support this approach.

Matt, Clint, and I have done some work defining “human activity” but it has already been criticized on at least one issue thread (for some ontology) as being too species specific.

I think it needs to be more general as you say. This would be very helpful in a number of areas for OMRSE and other ontologies we’re working on.

On Tue, Sep 27, 2022 at 2:59 PM Chris Mungall @.***> wrote:

Just upping this as it came up in @hoganwr <https://github.com/hoganwr 's ICBO talk. I think people are using GO hevavior to categorize processes that are not genetically determined.

Defining behavior is hard. See: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2760923/

I suggest the following as a pragmatic approach:

• COB: process
• COB(new): behavioral action or activity of an organism
• going for a walk
• taking a class
• aggressive actions or activities
• COB/GO: biological process: genetically programmed
• GO(+COB-new?):behavior genetically programmed
• GO:learning or memory
• GO:locomotory behavior
• GO:aggressive behavior

Here the two classes would be class-disjoint. They could be linked by parthood relation at the domain ontology level. E.g with @dosumis https://github.com/dosumis' example, the NCIT:going-for-a-walk activity may have a has-part to GO:walking/locomotory behavior. COB would be neutral about when this would be done.

This approach can be criticised in that it can be really hard to know the boundaries. There will always be debates as to the extent of genetic influence on behavior, and things vary by species. We may end up with a bit of concept duplication and artificial labels to distinguish terms in the two hierarchies.

But it's very pragmatic. If there is evidence for the involvement in genes then the concept can be added to GO, if there are annotations. Otherwise keep it outside.

There is definitely much to be improved. I think the labels for GO behaviors should probably be modified to make their restricted sense clear. But my proposal here can go ahead without being blocked on this.

I don't know what that means for NBO, but it's inactive, and not really a COB problem. We just need a solution for people who need "behaviors" that don't belong in GO.

— Reply to this email directly, view it on GitHub https://github.com/OBOFoundry/COB/issues/156#issuecomment-1259925815, or unsubscribe < https://github.com/notifications/unsubscribe-auth/AAJR55UCF3KGCS2OTBXMUO3WAM7Z5ANCNFSM4ZV3722Q

. You are receiving this because you were mentioned.Message ID: @.***>

— Reply to this email directly, view it on GitHub https://github.com/OBOFoundry/COB/issues/156#issuecomment-1259949013, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADJX2ISDFQQYRBQFAWERGKDWANCRDANCNFSM4ZV3722Q . You are receiving this because you are subscribed to this thread.Message ID: @.***>

-- Bjoern Peters Professor La Jolla Institute for Immunology 9420 Athena Circle La Jolla, CA 92037, USA Tel: 858/752-6914 Fax: 858/752-6987 http://www.liai.org/pages/faculty-peters

[hoganwr]

Agree. They are not the same. Certain compulsive behaviors are not programmed nor are they planned. They are closely related and teasing out the distinctions will be important.

On Tue, Sep 27, 2022 at 3:50 PM bpeters42 @.***> wrote:

I am all in favor of giving these processes a home in COB, but would like to see the connection to 'planned process' being worked out well. It seems that with organism behavior, there typically is an 'objective', even if there isn't a full scale plan, so that needs to be made compatible.

• Bjoern

On Tue, Sep 27, 2022 at 12:22 PM Bill Hogan @.***> wrote:

+1

I support this approach.

Matt, Clint, and I have done some work defining “human activity” but it has already been criticized on at least one issue thread (for some ontology) as being too species specific.

I think it needs to be more general as you say. This would be very helpful in a number of areas for OMRSE and other ontologies we’re working on.

On Tue, Sep 27, 2022 at 2:59 PM Chris Mungall @.***> wrote:

Just upping this as it came up in @hoganwr <https://github.com/hoganwr 's ICBO talk. I think people are using GO hevavior to categorize processes that are not genetically determined.

Defining behavior is hard. See: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2760923/

I suggest the following as a pragmatic approach:

• COB: process
• COB(new): behavioral action or activity of an organism
• going for a walk
• taking a class
• aggressive actions or activities
• COB/GO: biological process: genetically programmed
• GO(+COB-new?):behavior genetically programmed
• GO:learning or memory
• GO:locomotory behavior
• GO:aggressive behavior

Here the two classes would be class-disjoint. They could be linked by parthood relation at the domain ontology level. E.g with @dosumis https://github.com/dosumis' example, the NCIT:going-for-a-walk activity may have a has-part to GO:walking/locomotory behavior. COB would be neutral about when this would be done.

This approach can be criticised in that it can be really hard to know the boundaries. There will always be debates as to the extent of genetic influence on behavior, and things vary by species. We may end up with a bit of concept duplication and artificial labels to distinguish terms in the two hierarchies.

But it's very pragmatic. If there is evidence for the involvement in genes then the concept can be added to GO, if there are annotations. Otherwise keep it outside.

There is definitely much to be improved. I think the labels for GO behaviors should probably be modified to make their restricted sense clear. But my proposal here can go ahead without being blocked on this.

I don't know what that means for NBO, but it's inactive, and not really a COB problem. We just need a solution for people who need "behaviors" that don't belong in GO.

— Reply to this email directly, view it on GitHub <https://github.com/OBOFoundry/COB/issues/156#issuecomment-1259925815 , or unsubscribe <

https://github.com/notifications/unsubscribe-auth/AAJR55UCF3KGCS2OTBXMUO3WAM7Z5ANCNFSM4ZV3722Q

. You are receiving this because you were mentioned.Message ID: @.***>

— Reply to this email directly, view it on GitHub https://github.com/OBOFoundry/COB/issues/156#issuecomment-1259949013, or unsubscribe < https://github.com/notifications/unsubscribe-auth/ADJX2ISDFQQYRBQFAWERGKDWANCRDANCNFSM4ZV3722Q

. You are receiving this because you are subscribed to this thread.Message ID: @.***>

-- Bjoern Peters Professor La Jolla Institute for Immunology 9420 Athena Circle https://www.google.com/maps/search/9420+Athena+Circle+%0D%0D%0ALa+Jolla,+CA+92037,+USA?entry=gmail&source=g La Jolla, CA 92037, USA https://www.google.com/maps/search/9420+Athena+Circle+%0D%0D%0ALa+Jolla,+CA+92037,+USA?entry=gmail&source=g Tel: 858/752-6914 Fax: 858/752-6987 http://www.liai.org/pages/faculty-peters

— Reply to this email directly, view it on GitHub https://github.com/OBOFoundry/COB/issues/156#issuecomment-1259975782, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAJR55V3YS34ER5HBB3BQQDWANFWTANCNFSM4ZV3722Q . You are receiving this because you were mentioned.Message ID: @.***>

[dosumis]

It is very naive to think we can easily make simple distinctions between genetically programmed / reflexive vs learned behaviour. Even the simplest behaviours (classic reflexes) can be modified by experience. Even complex behaviours rely on evolved circuits structured according to genetically determined developmental programs.

Naming a term "behaviour - genetically programmed" will cause lots of well justified push back from biologists. (Actually - that's an understatement - it will make GO look ridiculous to most potential users working in neurobiology/behavior).

I think we can split out human, culturally-located(?) behaviour terms ("going to the gym", "smoking", "taking a class", "getting married"). But I'm very wary of doing anything more than that.

If GO wants to reject a wider range of terms than this, then these terms will need another home, and if they don't want these terms to live under GO process in a different namespace (e.g. NBO), then we need a different grouping term.

[dosumis]

would like to see the connection to 'planned process' being worked out well. It seems that with organism behavior, there typically is an 'objective', even if there isn't a full scale plan, so that needs to be made compatible.

I'm not convinced. For the OBI use case - distinguishing types of process in an experiment - 'planned process' totally makes sense. In the context of annotating behaviours in, say, a GWAS experiment, it does not.

Matt, Clint, and I have done some work defining “human activity” but it has already been criticised on at least one issue thread (for some ontology) as being too species specific.

Other species have culture too . Maybe:

'culturally-located activity' as a sibling term to behaviour?

[bpeters42]

The concept of objectives, plans and actions is very well defined far beyond OBI. Calling something "culturally-located(?) behaviour term" sounds super ridiculous to me, but clearly we are coming this from different perspectives. It would be useful if you could define what you mean; the OBI definitions are out there to review, and are supposed to be part of COB specifically because they have a broader scope than investigations. I am sure they can be improved, but not everything has to start and end in GO.

• Bjoern

On Wed, Sep 28, 2022 at 1:32 AM David Osumi-Sutherland < @.***> wrote:

would like to see the connection to 'planned process' being worked out well. It seems that with organism behavior, there typically is an 'objective', even if there isn't a full scale plan, so that needs to be made compatible.

I'm not convinced. For the OBI use case - distinguishing types of process in an experiment - 'planned process' totally makes sense. In the context of annotating behaviours in, say, a GWAS experiment, it does not.

Matt, Clint, and I have done some work defining “human activity” but it has already been criticised on at least one issue thread (for some ontology) as being too species specific.

Other species have culture too maybe:

'culturally-located activity' as a sibling term to behaviour?

— Reply to this email directly, view it on GitHub https://github.com/OBOFoundry/COB/issues/156#issuecomment-1260569181, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADJX2IRSHE7ZQ4DBQ6ZGZ33WAP7BRANCNFSM4ZV3722Q . You are receiving this because you commented.Message ID: @.***>

-- Bjoern Peters Professor La Jolla Institute for Immunology 9420 Athena Circle La Jolla, CA 92037, USA Tel: 858/752-6914 Fax: 858/752-6987 http://www.liai.org/pages/faculty-peters

[wdduncan]

I would stay away from using objective specification as a general differentia for defining behavior. You wade deeper into the murky idea that concretizations can be realizables (I still don't think that make sense). Plus, it seems (to me) to be asserting that every behavior an organism engages in has to have objective specification related to it, regardless of how loosely defined the objective specification may be. I am skeptical that frogs, flies, amoeba, and hydra have an objective specification related to every behavior they engage in.

[bpeters42]

My response was careless: I was intending to only refer to the 'human activities' part that Chris had outlined (human activities like walking, smoking, going to the gym, ...) and Bill responded to as suggesting they could go under planned process. I agree that there might be more broader parent terms for behaviors, genetically determined or not, and I don't have strong feelings how they would be modeled as long as we don't build a parallel overlapping set of terms to 'planned process' which has served us very well.

• Bjoern

On Thu, Sep 29, 2022 at 7:21 AM Bill Duncan @.***> wrote:

I would stay away from using objective specification as a general differentia for defining behavior. You wade deeper into the murky idea that concretizations can be realizables (I still don't think that make sense). Plus, it seems (to me) to be asserting that every behavior an organism engages in has to have objective specification related to it, regardless of how loosely defined the objective specification may be. I am skeptical that frogs, flies, amoeba, and hydra have an objective specification related to every behavior they engage in.

— Reply to this email directly, view it on GitHub https://github.com/OBOFoundry/COB/issues/156#issuecomment-1262356545, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADJX2IXQ7CIFHLWVOIWZIHLWAWQWHANCNFSM4ZV3722Q . You are receiving this because you commented.Message ID: @.***>

-- Bjoern Peters Professor La Jolla Institute for Immunology 9420 Athena Circle La Jolla, CA 92037, USA Tel: 858/752-6914 Fax: 858/752-6987 http://www.liai.org/pages/faculty-peters

[cmungall]

@dosumis:

Naming a term "behaviour - genetically programmed" will cause lots of well justified push back from biologists. (Actually - that's an understatement - it will make GO look ridiculous to most potential users working in neurobiology/behavior).

GO behavior: "The internally coordinated responses (actions or inactions) of animals (individuals or groups) to internal or external stimuli, via a mechanism that involves nervous system activity"

is currently a subclass of GO biological process, "A biological process represents a specific objective that the organism is genetically programmed to achieve."

I agree that this is problematic on a few levels but from the perspective of this issue, that is the definitions in GO. If we want to discuss changing definitions in GO, this should be done on the GO issue tracker.

[bpeters42]

So it seems like right now we have something like: 

• organismal process [made up]   
  • biological process [go]           - behavior [go]                    - planned process [cob]                

And I believe we are discussing variants such as this:

• organismal process [made up]   
  • biological process [go]
  • behavior [go]                    - planned process [cob]

to avoid that all 'behaviors inherit the axioms from GO: biological process, which imply genetic determination? I would be very much in favor of that.

[ddooley]

OBI and others are putting assays, calibration, data transformation etc. under planned process - and robots or devices may be executing them, so "planned process" would be peer to organismal process. Maybe a new term "organismal planned process" could be peer to "biological process" though.