sample

[obi-bot]

A Sample is a role which may be realised by an entity in a process where the entity is a subset of a population (whole or class)

Reported by: fgibson_sf

Original Ticket: "obi/obi-terms/32":https://sourceforge.net/p/obi/obi-terms/32

[obi-bot]

• assigned_to: nobody --> fostel

Original comment by: mcourtot

[obi-bot]

• assigned_to: fostel --> jenfostel

Original comment by: mcourtot

[obi-bot]

Logged In: YES user_id=1535944 Originator: NO

this may be synonymous with subject, defined as a population representative

Original comment by: jenfostel

[obi-bot]

• status: open --> pending

Original comment by: jenfostel

[obi-bot]

Logged In: YES user_id=1430184 Originator: YES

I think sample would be synonymous with subject, However the definition does not quite fit. The "in most cases...." seems rather vague for an explicit ontological statement of knowledge. This is the current definition of subject.

"Study subject is a Study participant role that is played by the objects of the study during the study timeline. In most cases, a study subject is considered to be a population representative"

I would suggest changing this to:

Study subject is a Study participant role that is played by the objects of the study during the study timeline. A study subject is a subset of a population (whole or class) which is assumed to be representative of the population.

I would be happy then to have sample be synonymous with subject.

Original comment by: fgibson_sf

[obi-bot]

• status: pending --> open

Original comment by: fgibson_sf

[obi-bot]

Logged In: YES user_id=138355 Originator: NO

I don't think sample and subject coincide, given Frank's comments, as we sometimes explicitly study mutants as unrepresentative. I would name this representative_sample_role so as not to have people get confused and have it be a subclass of subject_role.

Original comment by: alanruttenberg

[obi-bot]

Logged In: YES user_id=891804 Originator: NO

I think two things are being conflated in Frank's proposal and also in the original definition. The subject role and being representative of a population are two separate things. In cohort type experiments yes we do expect that things are representative of a population, in other experiments this is not the case. I would like to see sample_role retained only to specify that this is the object being studied and some other role used for cases where the 'representativeness is important'. In this case the subject is proxy for a population perhaps?

Original comment by: helenp

[obi-bot]

Logged In: YES user_id=1430184 Originator: YES

If you study a mutant - say a mouse, would you not be studying it because you want to draw conclusions about gene expression? The mutant is a sample (a subset) the population (whole or class) of all mice with that mutation, which can be used to infer/draw conclusions - accept/reject the null hypothesis about the mouse genome? Maybe its my world view, but I see science experiments as studying samples (subsets) to draw conclusions about the population/whole/class, i.e an organism. Maybe this is a very high level sweeping generalisation and we require more specific descriptions....

Helen's point of "the object being studied" introduces another angle and comes very close to what an analyte was being proposed to be. However, a mutant mouse in a experiment is a sample (a subset of all mutant mice) in the experiment it is also the object of study (analyte). But lets not bring up the analyte discussion again here :)

Original comment by: fgibson_sf

[obi-bot]

Logged In: YES user_id=1430184 Originator: YES

I have thought about this some more in order to try and gain a consensus. Maybe it would be more appropriate to talk about a sample with respect to statistics and in respect to materials, allowing for an object to be both a statistical sample and a sample of material, or only one if appropriate. I still believe all science experiments are carried out using a statistical sample of material.

So something like this - massaged into the appropriate role definition format.

Sample (statistics), the part (subset) of a population (whole or class) which is assumed to be representative of the population.

Sample (material), a part or subset or limited quantity of material, usually from a larger population (whole or class)

Original comment by: fgibson_sf

[obi-bot]

Logged In: YES user_id=1868269 Originator: NO

we have two terms for this one word.
the best fit is study subject: Study subject is a Study participant role that is played by the objects of the study during the study timeline. In most cases, a study subject is considered to be a population representative.

this is the current term that best fits the term sample as used here. i can add sample (statistical) as an alternative term if you would like. we had consdiered adding the term population representative, but could not distinguish it from subject.

to many people, sample can also mean a biological specimen A biological specimen is a study participant consisting of a section of tissue, sample of fluid, or other material removed from a subject during a study. this is not coming from a larger population (eg there is only one left lobe of a liver in the entire rat) although the rat itself is a population representative.

Original comment by: jenniferfostel

[obi-bot]

Logged In: YES user_id=1430184 Originator: YES

Hi Jennifer, what is the current status of this tracker, you seemed to have republished the original comment from which the discussion ensued?

Original comment by: fgibson_sf

[obi-bot]

• status: open --> closed

Original comment by: jenfostel

[obi-bot]

please dont close a tracker item without posting the solution - As far as I am aware the issues have not been addresses

Original comment by: fgibson_sf

[obi-bot]

• status: closed --> open

Original comment by: fgibson_sf

[obi-bot]

to move this it would be helpful to add the process where the entity is a subset of a population.

from the definition given this sounds different from the process of sampling a population, the output of which would be a population representative. we can add population representative role to OBI. for this it would be helpful to define the process of creating a population representative.

Original comment by: jenniferfostel

[obi-bot]

• status: open --> closed

Original comment by: bpeters42

[obi-bot]

This is closed, as it is superseded by current sample / specimen etc. role etc. discussion.

Original comment by: bpeters42

[obi-bot]

Here are the definitions based on which this was closed.

specimen role :=def a role borne by a material entity obtained through a specimen gathering process and that is realized by the process of gathering and potentially storing and using in an investigation

material sample role =def A material sample role is a specimen role borne by a material entity that is the output of a material sampling process.

material sampling process =def: A specimen gathering process with the objective to obtain a specimen that is representative of the input material entity

Original comment by: bpeters42