Re: [Update] Biomaterial Branch @ Fri, March 20, 8:30am – 9:30am
1 view
Skip to first unread message
Alan Ruttenberg
Mar 17, 2009, 12:39:13 AM3/17/09
to Bjoern Peters, OBI Biomaterial
ccing obi-biomaterial instead of frankenstein cc list.
I always like to look at http://www.golovchenko.org/cgi-bin/wnsearch?q=specimen
Note that it says that the *word* specimen isn't univocal. We are
going to have to deal with that, probably by not using the bare term
in the ontology.
Note the wikipedia entry takes a bit of both:
http://en.wikipedia.org/wiki/Specimen
Note the the sample sense would need to be relational somehow - two
things needed to describe it - the thing taken, and the larger thing
from which it is considered representative.
On Mon, Mar 16, 2009 at 9:53 PM, Bjoern Peters <bpe...@liai.org> wrote:
> A few questions to prepare that discusssions:
> What material entities can be a specimen? (bear the specimen role)
> - clearly: former part of organism (e.g. blood sample)
>
> But what about:
> - single organism (mouse caught from a wild population)
Wikipedia says yes.
> - population of organisms (e.g. sample of fungal growth on a tree)
Wikipedia says yes.
> - cell culture (derived from blood from a patient)
For the sense of "sample", yes. Otherwise not.
> - is a HeLa cell culture a specimen?
I don't think so. But I can take a specimen of a HeLa cell (in the
sense of sample/representative)
Just going on what we can.
I think there are at least two things here.
1) is the representativeness
2) is the result of the taking process
-Alan
>
> - Bjoern
>
> Melanie Courtot wrote:
>>
>> Hi all,
>> An early reminder for next week's biomaterial call. We would like to go on
>> discussing the specimen (ex-EOO) case, and would appreciate if all were
>> there to take part in the discussion. Hopefully a joint operation between
>> biomaterial, process and role will finally manage to sort that one out :)
>> Thanks,
>> Melanie
>
>
Melanie Courtot
Mar 17, 2009, 1:08:36 AM3/17/09
to obi-bio...@googlegroups.com, Bjoern Peters
On 16-Mar-09, at 9:39 PM, Alan Ruttenberg wrote:
>
> ccing obi-biomaterial instead of frankenstein cc list.
>
> I always like to look at http://www.golovchenko.org/cgi-bin/wnsearch?q=specimen
>
> Note that it says that the *word* specimen isn't univocal. We are
> going to have to deal with that, probably by not using the bare term
> in the ontology.
>
> Note the wikipedia entry takes a bit of both:
> http://en.wikipedia.org/wiki/Specimen
>
> Note the the sample sense would need to be relational somehow - two
> things needed to describe it - the thing taken, and the larger thing
> from which it is considered representative.
seems to say sample = specimen and core, where sample = element
representative of its class, includes specimen for biological samples
(blood, urine) and core for mineral samples (soil,rock)
I think we are making that distinction when saying specimen
"derives_from" organism.
Can we start with the assumption that in our case, where we are
talking about entities derived from organism, sample and specimen are
synonyms?
>
>
> On Mon, Mar 16, 2009 at 9:53 PM, Bjoern Peters <bpe...@liai.org>
> wrote:
>> A few questions to prepare that discusssions:
>> What material entities can be a specimen? (bear the specimen role)
>> - clearly: former part of organism (e.g. blood sample)
>>
>> But what about:
>> - single organism (mouse caught from a wild population)
> Wikipedia says yes.
for whole organisms, instead using subjects.
I think we may have to make a choice for the general use and then use
alternative terms for other communities.
>
>> - population of organisms (e.g. sample of fungal growth on a tree)
> Wikipedia says yes.
>
>> - cell culture (derived from blood from a patient)
> For the sense of "sample", yes. Otherwise not.
representative of the blood they are coming from.
>
>> - is a HeLa cell culture a specimen?
> I don't think so. But I can take a specimen of a HeLa cell (in the
> sense of sample/representative)
>
>
> Just going on what we can.
> I think there are at least two things here.
> 1) is the representativeness
> 2) is the result of the taking process
the first one being the taking process. As long as the
representativeness quality is kept I would think it is still a
specimen. Maybe using representative as relational quality towards
organism(e.g mouse)/specimen (e.g. blood)/population (e.g fungal
growth)?
(note: we used the example of an histological preparation during the
last call: I take a slice of liver (taking from organism process). I
stain the tissue with a dye (staining process) at the end of the
second process the liver tissue is still a specimen.)
Melanie
>
>
> -Alan
>>
>> - Bjoern
>>
>> Melanie Courtot wrote:
>>>
>>> Hi all,
>>> An early reminder for next week's biomaterial call. We would like
>>> to go on
>>> discussing the specimen (ex-EOO) case, and would appreciate if all
>>> were
>>> there to take part in the discussion. Hopefully a joint operation
>>> between
>>> biomaterial, process and role will finally manage to sort that one
>>> out :)
>>> Thanks,
>>> Melanie
>>
>>
>
> >
---
>
> -Alan
>>
>> - Bjoern
>>
>> Melanie Courtot wrote:
>>>
>>> Hi all,
>>> An early reminder for next week's biomaterial call. We would like
>>> to go on
>>> discussing the specimen (ex-EOO) case, and would appreciate if all
>>> were
>>> there to take part in the discussion. Hopefully a joint operation
>>> between
>>> biomaterial, process and role will finally manage to sort that one
>>> out :)
>>> Thanks,
>>> Melanie
>>
>>
>
> >
Mélanie Courtot
TFL- BCCRC
675 West 10th Avenue
Vancouver, BC
V5Z 1L3, Canada
Frank Gibson
Mar 17, 2009, 6:06:18 AM3/17/09
to obi-bio...@googlegroups.com, Bjoern Peters
There has been a long thread on this a while ago, which seems to have
been forgoten about. I tried to find it in the archives, but it seems
either not to have been recorded at all, or in bits across numerous
archives. I am going to re-send it to the people listed here, rather
than spamming the list.
As far as Jennifers community wanting a different deffinition of
specimen, I think this is another case of confusing the general term,
with a specfic use of the terms. the role drug is another example of
this. I would argue that all chemicals that can be called drugs are
not all FDA approved, as per the definition in the file. I raised this
issue at the time and presented alternatives.
Specimens and sample are synonyms for the same thing. I have carried
out experiments where my specimen is a particular type of plastic
petri dish, it is not only restricted to organism deriviatives/parts
Frank
--
Frank Gibson, PhD
http://peanutbutter.wordpress.com/
been forgoten about. I tried to find it in the archives, but it seems
either not to have been recorded at all, or in bits across numerous
archives. I am going to re-send it to the people listed here, rather
than spamming the list.
As far as Jennifers community wanting a different deffinition of
specimen, I think this is another case of confusing the general term,
with a specfic use of the terms. the role drug is another example of
this. I would argue that all chemicals that can be called drugs are
not all FDA approved, as per the definition in the file. I raised this
issue at the time and presented alternatives.
Specimens and sample are synonyms for the same thing. I have carried
out experiments where my specimen is a particular type of plastic
petri dish, it is not only restricted to organism deriviatives/parts
Frank
Frank Gibson, PhD
http://peanutbutter.wordpress.com/
James Malone
Mar 17, 2009, 8:08:39 AM3/17/09
to obi-bio...@googlegroups.com, James Malone
Here are the definitions I would consider:
sample (syn specimen): A material entity has_role sample is an example regarded as typical of its class.
NB: when I say 'sample' above I really mean 'material sample', we would need to add statistical sample to deal with non-material stats samples.
sample collection (syn sampling): A process in which a material entity is acquired from another material entity from which it is regarded as typical of that class (has_role sample) .
material collection: Process of collecting material, subclasses environmental matter collection and so on.
The drug role thing aside (which I agree with, I don't see FDA approval as nec to this definition) I can also see the following use cases being relevant:
1. My human is a sample of a whole population of humans (such as ethnic groups in GWAS)
2. My blood is a sample taken from a mouse but and that is representative of all the blood in the mouse
3. My original sample is a slice of liver, then I stain it in order to do some further analysis and I still consider this my sample (despite the extra processing) (from Bjoern's original set).
4. My petri dish is a sample of a population of petri dishes (from Frank).
So for #1, we say that the sample is considered typical of its class. This seems different for Jeniffer's community. My question would be why? When you do a test on a human as a subject are you considering them as a sample of a population (such as an ethnic group) or are you just considering bits of them in isolation, in which case the bits are the samples of the larger class of things they are taken from and then the human is the material it is derived from (note: blood is a sample of all blood and not a sample of a human; it is not representative of the whole human, which I think is the point Frank is making). I think this still fits our definition here.
For #2, this also fits our definition, again I think this is not controversial, other than to say it is a sample of all_that_mouses_blood and is derived_from mouse. Again looking at Alan's earlier definition "typical of its class", blood being the class here. I think we need to avoid being loose here. A blood sample is a sample of blood, not a sample of human.
For #3, we can consider Alan's second definition (A bit of tissue or blood or urine that is taken for diagnostic purposes; "they collected a urine specimen for urinalysis") this defines something different - this defines the process of collection of the sample and I think this conflicts with what we should be saying. We presently agree with this and say that the output of environmental matter collection is a specimen. However, it is not specified that it is representative of anything. I would suggest the output of an envrinonmental matter collection processes is envrinonmental matter. More specifically, the output of a process 'specimen collection' is a specimen and here we would explicitly say that the output specimen is representative of the larger class of things we are sampling.
I also think the point Melanie makes about multiple 'chains' of processes is a valid one; I would assume it is still a sample after it's been stained, although it presumably also gains other qualities along the way through this chain so we can keep track that it's a sample but it's also stained, chopped and fried with liver. Delicious.
For #4, we can still use the definition for case 1 above. Still fits assuming we make our restrictions on material entity level and make definition of sample "'something representative of its class".
James
sample (syn specimen): A material entity has_role sample is an example regarded as typical of its class.
NB: when I say 'sample' above I really mean 'material sample', we would need to add statistical sample to deal with non-material stats samples.
sample collection (syn sampling): A process in which a material entity is acquired from another material entity from which it is regarded as typical of that class (has_role sample) .
material collection: Process of collecting material, subclasses environmental matter collection and so on.
The drug role thing aside (which I agree with, I don't see FDA approval as nec to this definition) I can also see the following use cases being relevant:
1. My human is a sample of a whole population of humans (such as ethnic groups in GWAS)
2. My blood is a sample taken from a mouse but and that is representative of all the blood in the mouse
3. My original sample is a slice of liver, then I stain it in order to do some further analysis and I still consider this my sample (despite the extra processing) (from Bjoern's original set).
4. My petri dish is a sample of a population of petri dishes (from Frank).
So for #1, we say that the sample is considered typical of its class. This seems different for Jeniffer's community. My question would be why? When you do a test on a human as a subject are you considering them as a sample of a population (such as an ethnic group) or are you just considering bits of them in isolation, in which case the bits are the samples of the larger class of things they are taken from and then the human is the material it is derived from (note: blood is a sample of all blood and not a sample of a human; it is not representative of the whole human, which I think is the point Frank is making). I think this still fits our definition here.
For #2, this also fits our definition, again I think this is not controversial, other than to say it is a sample of all_that_mouses_blood and is derived_from mouse. Again looking at Alan's earlier definition "typical of its class", blood being the class here. I think we need to avoid being loose here. A blood sample is a sample of blood, not a sample of human.
For #3, we can consider Alan's second definition (A bit of tissue or blood or urine that is taken for diagnostic purposes; "they collected a urine specimen for urinalysis") this defines something different - this defines the process of collection of the sample and I think this conflicts with what we should be saying. We presently agree with this and say that the output of environmental matter collection is a specimen. However, it is not specified that it is representative of anything. I would suggest the output of an envrinonmental matter collection processes is envrinonmental matter. More specifically, the output of a process 'specimen collection' is a specimen and here we would explicitly say that the output specimen is representative of the larger class of things we are sampling.
I also think the point Melanie makes about multiple 'chains' of processes is a valid one; I would assume it is still a sample after it's been stained, although it presumably also gains other qualities along the way through this chain so we can keep track that it's a sample but it's also stained, chopped and fried with liver. Delicious.
For #4, we can still use the definition for case 1 above. Still fits assuming we make our restrictions on material entity level and make definition of sample "'something representative of its class".
James
-- European Bioinformatics Institute, Wellcome Trust Genome Campus, Hinxton, Cambridge, CB10 1SD, United Kingdom Tel: + 44 (0) 1223 494 676 Fax: + 44 (0) 1223 492 468
Bjoern Peters
Mar 17, 2009, 10:13:25 AM3/17/09
to Frank Gibson, obi-bio...@googlegroups.com, OBI Developers
That discussion is not forgotten, but there was never a consensus
resolution, as was evident at the workshop when subject / sample /
specimen were discussed. The goal of re-raising this discussion here was
to see specifically how it relates to 'entity of organismal origin'
(EOO). It was off-list to work on a coherent proposal before presenting
it, to avoid exactly what is happening now, namely the re-discussion of
the general case of sample / specimen, rather than specific use for EOO
case. If we can reach agreement quick, great. If not, I would want to
cut that discussion off, and focus on EOO only.
- Bjoern
resolution, as was evident at the workshop when subject / sample /
specimen were discussed. The goal of re-raising this discussion here was
to see specifically how it relates to 'entity of organismal origin'
(EOO). It was off-list to work on a coherent proposal before presenting
it, to avoid exactly what is happening now, namely the re-discussion of
the general case of sample / specimen, rather than specific use for EOO
case. If we can reach agreement quick, great. If not, I would want to
cut that discussion off, and focus on EOO only.
- Bjoern
Frank Gibson
Mar 17, 2009, 11:15:26 AM3/17/09
to Bjoern Peters, obi-bio...@googlegroups.com, OBI Developers
samples(syn:specimens) may or may not be derived from EOO. All EOO are
not samples(specimens). A material_sample(specimen) is representative
of a material class(whole)
Frank.
not samples(specimens). A material_sample(specimen) is representative
of a material class(whole)
Frank.
James Malone
Mar 17, 2009, 11:19:55 AM3/17/09
to obi-bio...@googlegroups.com, Bjoern Peters, OBI Developers
Agreed, the only thing EOO should say is that its origin was an
organism, i.e. material entity derived from organism. I think a few
calls ago we decided to call this specimen but I think this was wrong.
We need to separate out things that are samples and things that were
derived from organisms as they're different.
James
organism, i.e. material entity derived from organism. I think a few
calls ago we decided to call this specimen but I think this was wrong.
We need to separate out things that are samples and things that were
derived from organisms as they're different.
James
![Fostel, Jennifer (NIH/NIEHS) [C]'s profile photo](http://lh3.googleusercontent.com/a-/ALV-UjWaRNmw5a_d1c3gde_voKpBx2FaCMfil1LSXKzMs4AL96z24w=s40-c)
Fostel, Jennifer (NIH/NIEHS) [C]
Mar 17, 2009, 11:31:19 AM3/17/09
to James Malone, obi-bio...@googlegroups.com, OBI Developers
agree with you about the word sample, which was why i recommended using specimen for EOO
it seems from your message that you are using specimen and sample interchangeably, while i think they mean different things. i will stop arguing at this point.
we need a word for EOO. i withdraw the idea of using specimen for this if some think sample = specimen. if there is no consensus here let us use EOO and accept derision.
we need a word to represent sample = specimen != EOO. please suggest one. we cannot use sample or specimen as it causes confusion and lengthy discussion.
...jennifer
------------------------------------------------------------------------------
Apps built with the Adobe(R) Flex(R) framework and Flex Builder(TM) are powering Web 2.0 with engaging, cross-platform capabilities. Quickly and easily build your RIAs with Flex Builder, the Eclipse(TM)based development software that enables intelligent coding and step-through debugging.
Download the free 60 day trial. http://p.sf.net/sfu/www-adobe-com _______________________________________________
Obi-devel mailing list
Obi-...@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/obi-devel
it seems from your message that you are using specimen and sample interchangeably, while i think they mean different things. i will stop arguing at this point.
we need a word for EOO. i withdraw the idea of using specimen for this if some think sample = specimen. if there is no consensus here let us use EOO and accept derision.
we need a word to represent sample = specimen != EOO. please suggest one. we cannot use sample or specimen as it causes confusion and lengthy discussion.
...jennifer
Apps built with the Adobe(R) Flex(R) framework and Flex Builder(TM) are powering Web 2.0 with engaging, cross-platform capabilities. Quickly and easily build your RIAs with Flex Builder, the Eclipse(TM)based development software that enables intelligent coding and step-through debugging.
Download the free 60 day trial. http://p.sf.net/sfu/www-adobe-com _______________________________________________
Obi-devel mailing list
Obi-...@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/obi-devel
James Malone
Mar 17, 2009, 11:34:28 AM3/17/09
to Fostel, Jennifer (NIH/NIEHS) [C], obi-bio...@googlegroups.com, OBI Developers
"material derived from organism" seems fine to me. Sorry Jen, I wasn't
overly clear in my email, I was using sample/specimen as synonyms as
you correctly pointed out (I sent a lenghty email on the list
explaining what I mean in a renamed thread).
Cheers!
James
overly clear in my email, I was using sample/specimen as synonyms as
you correctly pointed out (I sent a lenghty email on the list
explaining what I mean in a renamed thread).
Cheers!
James
Frank Gibson
Mar 17, 2009, 11:56:42 AM3/17/09
to obi-bio...@googlegroups.com, James Malone, OBI Developers
> agree with you about the word sample, which was why i recommended using specimen for EOO
> it seems from your message that you are using specimen and sample interchangeably, while i think they mean different things. i will stop arguing at this point.
sample and specimen are synonyms for the class defined as a material
> it seems from your message that you are using specimen and sample interchangeably, while i think they mean different things. i will stop arguing at this point.
that is representative of a whole(class)
>
> we need a word for EOO. i withdraw the idea of using specimen for this if some think sample = specimen. if there is no consensus here let us use EOO and accept derision.
assumed should be material is_derived_from some organism. The
additional restriction is output of specimen_creation seems strange,
but I see the editor not also highlights this. EOO can be a bi-product
of any process. I dont see any issue with the EOO lable for this class
>
> we need a word to represent sample = specimen != EOO. please suggest one. we cannot use sample or specimen as it causes confusion and lengthy discussion.
a sample_role is a role borne by a material that is assumed to be
representative of a population (class)
synonoms: specimen
EOO is a material that is derived from some organsim. This class is
just there to aid classification, it does not add new information, its
just a query. The label is fine or you could try
organism-derived-material, or material_derived_from_organism, its a
label, who cares!
Please refer to the issue tracker for the proposal presented over a
year ago [ 1901511 ] for futher details
Frank
James Malone
Mar 17, 2009, 12:11:52 PM3/17/09
to obi-bio...@googlegroups.com, OBI Developers
So I think to usefully get anywhere with this we should remove the word
'specimen' from our definitions/labels as the definition does not appear
to be separable in a discerning way for people. I think sample is less
controversial and I think the intent of EOO is reasonably uncontentious
(i.e. material derives_from organism). Can we agree on those two?
James
'specimen' from our definitions/labels as the definition does not appear
to be separable in a discerning way for people. I think sample is less
controversial and I think the intent of EOO is reasonably uncontentious
(i.e. material derives_from organism). Can we agree on those two?
James
Frank Gibson
Mar 17, 2009, 12:24:17 PM3/17/09
to obi-bio...@googlegroups.com, OBI Developers
On Tue, Mar 17, 2009 at 4:11 PM, James Malone <mal...@ebi.ac.uk> wrote:
>
> So I think to usefully get anywhere with this we should remove the word
> 'specimen' from our definitions/labels as the definition does not appear
> to be separable in a discerning way for people. I think sample is less
> controversial and I think the intent of EOO is reasonably uncontentious
> (i.e. material derives_from organism). Can we agree on those two?
I have no problem with this
>
> So I think to usefully get anywhere with this we should remove the word
> 'specimen' from our definitions/labels as the definition does not appear
> to be separable in a discerning way for people. I think sample is less
> controversial and I think the intent of EOO is reasonably uncontentious
> (i.e. material derives_from organism). Can we agree on those two?
Frank
Fostel, Jennifer (NIH/NIEHS) [C]
Mar 17, 2009, 12:27:55 PM3/17/09
to James Malone, obi-bio...@googlegroups.com, OBI Developers
we need to take out both specimen and sample; we are using them inconsistently and this is driving lengthy discussion.
we should propose a new term to represent the concept you currently represent using the words specimen and sample interchangeably and add that to OBI.
we need somehow to link specimen to EOO using alternative term tag (oops, we just elimnated this from metadata though!)
we should propose a new term to represent the concept you currently represent using the words specimen and sample interchangeably and add that to OBI.
we need somehow to link specimen to EOO using alternative term tag (oops, we just elimnated this from metadata though!)
Fostel, Jennifer (NIH/NIEHS) [C]
Mar 17, 2009, 12:28:46 PM3/17/09
to Frank Gibson, obi-bio...@googlegroups.com, OBI Developers
see my note -- i DO have a problem with this.
since you have a problem with the solution proposed by the PPA branch, and at least one of us takes issue with this solution we need a middle ground.
since you have a problem with the solution proposed by the PPA branch, and at least one of us takes issue with this solution we need a middle ground.
James Malone
Mar 17, 2009, 12:35:16 PM3/17/09
to Fostel, Jennifer (NIH/NIEHS) [C], Frank Gibson, obi-bio...@googlegroups.com, OBI Developers
So, labels aside, two things we seem to need:
1. A material that was derived from an organism.
2. A material that is representative of the type of thing it is and
which was taken from some material.
Can we agree on those two?
Jm
1. A material that was derived from an organism.
2. A material that is representative of the type of thing it is and
which was taken from some material.
Can we agree on those two?
Fostel, Jennifer (NIH/NIEHS) [C]
Mar 17, 2009, 2:40:35 PM3/17/09
to James Malone, Frank Gibson, obi-bio...@googlegroups.com, OBI Developers
i agree with #1.
how do we specify 2? nothing is truly a population representative in the real world, we are all indivuals. no studies are done on things not deemed representative of the type of thing that it is, so why specify. what are some use cases?
how do we specify 2? nothing is truly a population representative in the real world, we are all indivuals. no studies are done on things not deemed representative of the type of thing that it is, so why specify. what are some use cases?
Fostel, Jennifer (NIH/NIEHS) [C]
Mar 17, 2009, 3:09:57 PM3/17/09
to Bjoern Peters, obi-bio...@googlegroups.com, OBI Developers
i agree with you, and sadly we have lost our short cut name for the term you list below.
talk with you friday.
just checking your times are PDT.
...jennifer
-----Original Message-----
From: Bjoern Peters [mailto:bpe...@liai.org]
Sent: Tuesday, March 17, 2009 3:02 PM
To: obi-bio...@googlegroups.com
Cc: OBI Developers
Subject: Re: [Obi-devel] [obi-biomaterial] Re: [Update] Biomaterial Branch @ Fri, March 20, 8:30am - 9:30am
I will restate what I wrote earlier today. Instead of trying to work out the general #1 + #2, I thought we were trying to deal with a more specific class 'former organism part that is created/collected with the goal to be able to use it in an experiment in which it retains relevant properties it had as part of the source organism'.
This was based on Alan's approach, which has served us well in the past, to stop arguments at a high level by focusing on specifics instead.
Todays email thread validates to me that this is the approach we should take. Once we have agreement on this, we can revisit the higher level.
- Bjoern
Bjoern Peters
Assistant Member
La Jolla Institute for Allergy and Immunology 9420 Athena Circle La Jolla, CA 92037, USA
Tel: 858/752-6914
Fax: 858/752-6987
http://www.liai.org/pages/faculty-peters
------------------------------------------------------------------------------
talk with you friday.
just checking your times are PDT.
...jennifer
-----Original Message-----
From: Bjoern Peters [mailto:bpe...@liai.org]
Sent: Tuesday, March 17, 2009 3:02 PM
To: obi-bio...@googlegroups.com
Cc: OBI Developers
Subject: Re: [Obi-devel] [obi-biomaterial] Re: [Update] Biomaterial Branch @ Fri, March 20, 8:30am - 9:30am
This was based on Alan's approach, which has served us well in the past, to stop arguments at a high level by focusing on specifics instead.
Todays email thread validates to me that this is the approach we should take. Once we have agreement on this, we can revisit the higher level.
- Bjoern
Assistant Member
La Jolla Institute for Allergy and Immunology 9420 Athena Circle La Jolla, CA 92037, USA
Tel: 858/752-6914
Fax: 858/752-6987
http://www.liai.org/pages/faculty-peters
------------------------------------------------------------------------------
Bjoern Peters
Mar 17, 2009, 3:02:15 PM3/17/09
to obi-bio...@googlegroups.com, James Malone, Frank Gibson, OBI Developers
I will restate what I wrote earlier today. Instead of trying to work out
the general #1 + #2, I thought we were trying to deal with a more
specific class
'former organism part that is created/collected with the goal to be able
to use it in an experiment in which it retains relevant properties it
had as part of the source organism'.
This was based on Alan's approach, which has served us well in the past,
to stop arguments at a high level by focusing on specifics instead.
Todays email thread validates to me that this is the approach we should
take. Once we have agreement on this, we can revisit the higher level.
- Bjoern
the general #1 + #2, I thought we were trying to deal with a more
specific class
'former organism part that is created/collected with the goal to be able
to use it in an experiment in which it retains relevant properties it
had as part of the source organism'.
This was based on Alan's approach, which has served us well in the past,
to stop arguments at a high level by focusing on specifics instead.
Todays email thread validates to me that this is the approach we should
take. Once we have agreement on this, we can revisit the higher level.
- Bjoern
James Malone
Mar 17, 2009, 3:53:02 PM3/17/09
to obi-bio...@googlegroups.com, OBI Developers
> i agree with you, and sadly we have lost our short cut name for the term
> you list below.
Because we can't all agree on a single definition for 'specimen' that
> you list below.
matches the rest of the world's so therefore it was wrong, I am sorry to
say (I would much rather use specimen then EOO, believe me).
> I will restate what I wrote earlier today. Instead of trying to work out
> the general #1 + #2, I thought we were trying to deal with a more specific
> class
using Alan's approach of discussing specific components of the class.
There are (at least) 2 properties this class you describe below has. #1
the objects are former organism part, #2 they are collected for use in an
experiment in which it retains relevant properties it had as part of the
source organism. These parts are the definitions I have been trying to get
at with my definitions of 'sample' and 'material derived from organism'.
Clearly I am failing so I'm happy to back out until Friday. I have not
followed all the previous discussions on this so my contribution is
perhaps redundant.
James
EMBL Outstation - Hinxton,
European Bioinformatics Institute,
Wellcome Trust Genome Campus,
Hinxton,
Cambridge, CB10 1SD,
United Kingdom
Wellcome Trust Genome Campus,
Hinxton,
Cambridge, CB10 1SD,
United Kingdom
Frank Gibson
Mar 17, 2009, 5:08:42 PM3/17/09
to obi-bio...@googlegroups.com, OBI Developers
There seems to be alot of effort devoted to a query as opposed to
building the ontology. EOO is just a query, it adds no new knowledge
to the ontology.
I am afraid Bjoern I disagree with your perception. I am going to say
I am more annoyed than you :) - because I have documented proposals
for sample and specimen within the issue tracker over a year ago, that
continually refuses to be addressed.
In addition, this discussion has highlighted the the definition of the
derives_from relation is fundamentally broken and that has_grain
relation now conflicts with it., so it has been useful.
I would suggest you stop worrry about these high level query classes,
and focus on representing a single scientific experiment which OBI
still can not do.
I will re-state the solution to this issue.
a sample is a material that bears the role sample_role
a sample_role is a role borne by a material that is assumed to be
representative of a population (class)
synonoms: specimen
EOO is a material that is derived from some organsim. This class is
just there to aid classification, it does not add new information, its
just a query. The label is fine or you could try
organism-derived-material, or material_derived_from_organism, its a
label, who cares!
Frank
building the ontology. EOO is just a query, it adds no new knowledge
to the ontology.
I am afraid Bjoern I disagree with your perception. I am going to say
I am more annoyed than you :) - because I have documented proposals
for sample and specimen within the issue tracker over a year ago, that
continually refuses to be addressed.
In addition, this discussion has highlighted the the definition of the
derives_from relation is fundamentally broken and that has_grain
relation now conflicts with it., so it has been useful.
I would suggest you stop worrry about these high level query classes,
and focus on representing a single scientific experiment which OBI
still can not do.
I will re-state the solution to this issue.
a sample is a material that bears the role sample_role
a sample_role is a role borne by a material that is assumed to be
representative of a population (class)
synonoms: specimen
EOO is a material that is derived from some organsim. This class is
just there to aid classification, it does not add new information, its
just a query. The label is fine or you could try
organism-derived-material, or material_derived_from_organism, its a
label, who cares!
Fostel, Jennifer (NIH/NIEHS) [C]
Mar 18, 2009, 9:17:18 AM3/18/09
to Frank Gibson, obi-bio...@googlegroups.com, OBI Developers
we have study subject role in OBI, which inheres in the object of a study and is realized through the process of studying it. it is assumed to be a population representative but we permit the possibility of case studies about individuals so this is not a requirement at the moment. in order to define this role we needed to define study design, which is in essence the description of the experiment, but we cannot use the word experiment since it has specific meaning in MAGE.
can you give an example of a sample_role that is not the object of an experiment and still within scope of OBI? in what process is it realized? if we can differentiate this properly from study_subject_role we can consider adding it to the branch as a sibling.
this would help move us along.
we welcome input in the role discussions. please join us. the tracker and e-mail have not proven optimal venues for discussion.
thanks!
...jennifer
can you give an example of a sample_role that is not the object of an experiment and still within scope of OBI? in what process is it realized? if we can differentiate this properly from study_subject_role we can consider adding it to the branch as a sibling.
this would help move us along.
we welcome input in the role discussions. please join us. the tracker and e-mail have not proven optimal venues for discussion.
thanks!
...jennifer
Reply all
Reply to author
Forward
0 new messages