For Biobank data annotation, we need a couple of new terms of collection of specimen.
Sample family or specimen family
Definition: A collection of specimens that is the output of the material processing and input of the process is the same specimen. For example, aliquoting one specimen to multiple tubes and the collection of the aliquoted specimens are sample family.
Human Sample set or human specimen set
Definition: A collection of specimens that collect from one specific person at one specific time point. For example, on Sept 6, 2015, 3 tubes of blood specimen, 1 tube of urine specimen were collected from patient 1. Then 3 blood specimen and 1 urine specimen called sample set.
This term may be more general like: A collection of specimens that collect from one source at a specific time point with label 'sample set' or 'specimen set'.
The parent term 'collection of specimen' also need to be added.
Definition: An object_aggregate that has as parts two or more specimens.
object_aggregate and (has member only specimen) and (has member min 2 specimen)
It would be nice to have logical axioms associated with the terms. Current I don't have good one and would like to discuss it in the OBI call.
Ran out of time on OBI call, so adding my comment here:
1. Human sample set- should we include the "For example..." as part of the curator note rather than the definition.
*The reason is that human sample set might not include 3 tubes of blood and 1 of urine for every instance, therefore I think it should not be part of the definition (but still important to help understand the concept).
Last edit: Sagar 2015-10-26
Sagar, Thanks for your comments. That information will be put in the 'example of usage'.
Comments from Bjoern. Add 'specimen set collection process', a subClassOf 'specimen collection process' with output, 'specimen set'.
In addition, try to avoid using 'object aggregate' since there is granularity issue. We need to consider what is object first when use 'object aggregate'. The object will be changed at different granularity levels.
Last edit: jzheng 2015-10-29
Based on the discussion in OBI dev call on 10/26/2015. Here are the terms for adding to OBI.
Term: collection of specimens
Definition: A material entity that has part two or more specimens.
Logical axiom (eq): 'material entity' and ('has member' only specimen) and ('has member' min 2 specimen)
Example of usage: PBMC collected from multiple donors over a course of study
Term: specimen family
alternative label: sample family
Definition: A collection of specimens derived from one common specimen via one or more material processing process(es).
Logical axiom: collection of specimen and derived from some (specimen and has_role some 'family source specimen role')
Example of usage: Aliquoting one specimen to multiple tubes and the collection of the aliquoted specimens are a specimen family. Slicing a tissue specimen into multiple sections for microscopy.
Term: family source specimen role
Definition: a specimen role that inheres in a specimen that through various material processing results in multiple downstream specimens.
Editor note: create role to enable axiom in specimen family
Term: specimen set collection process
Definition: A specimen collection process that generates multiple specimens from one source (e.g. one organism) during a time period which for the purpose of the study can be considered to be taken at the same sampling time.
Logical axiom: has_part min 2 specimen collection process and has_specified_output some 'specimen set'
Example of usage: Collection both blood and urine specimens in one clinical visit; Taking out liver and brain specimens during an autopsy
Term: specimen set
Definition: A collection of specimens that is collected from one source (e.g. one organism) during a time period which for the purpose of the study can be considered to be taken at the same sampling time.
Logical axiom: is_specified_output_of some 'specimen set collection process'
Example of usage: both blood and urine specimens collected in one clinical visit; liver and brain specimens taken during an autopsy
Term: human specimen set
Definition: A specimen set that is collected from one person during a time period which for the purpose of the study can be considered to be taken at the same sampling time.
Logical axiom: 'specimen set' and derives from' some 'Homo sapiens'
Definitions and logical axioms updated during the OBI call 2015-11-02
Last edit: jzheng 2015-11-02
I'm glad to see the new definition for a collection of specimens. That will work well for BCO.
I am proposing alternative primary labels to make them more expressive over set / family
collection of specimens
- specimens derived from shared ancestor (above: specimen family)
- specimens collected in one encounter (above: specimen set)
- specimens collected longitudinally (not above, but would be a natural addition)
Bjoern
Last edit: jzheng 2015-11-04
Proposed term definition: specimens collected longitudinally
A collection of specimens that was derived from the same source material entity at different time points in order to observe changes in that entity.
I have doubts regarding the 'family source specimen role'. It seems to me that this role in the current representation is realized by the process that actually would create the role (creating aliquots).
I am not sure that what we try to capture here is a realizable entity. I will think about it and try to provide a solution by Monday Nov 16.
Following terms were added into OBI by Chris on Nov 17, 2015
collection of specimens
- specimens derived from shared ancestor
- specimens collected in one encounter
- human specimen set
- specimens collected longitudinally
Here is my proposal for dealing with specimen family and most importantly with family source specimen:
Term: specimen family creation
Definition: A specimen collection process that has as its input a specimen and as its output a collection of specimens.
Logical axiom: collection of specimen and has_specified_input some specimen and has_specified_output some specimen set
Example of usage: Aliquoting one specimen to multiple tubes, slicing a tissue specimen into multiple sections.
Term: family source specimen
Definition: A specimen that is the specified input to a specimen family creation.
Logical axiom: specimen and is_specified_input_to specimen family creation
Term: specimen family
alternative label: sample family
Definition: A collection of specimens that is the outcome of a specimen family creation .
Logical axiom: collection of specimen and derived from some (specimen and has_role some 'family source specimen role')
The specimen family axiom is wrong. It needs to be: collection of specimen and is_specified_output_of some specimen family creation
The output of a specimen collection process is specimen. During this process some material entity gain specimen role. So, specimen cannot be the input of this process.
Term: specimen family creation
Propose to change the definition to:
A planned process that has as its input a specimen and as its output a collection of specimens through one or more material processing process(s).
Propose to add logical axiom:
has part some 'material processing'
Discussed on obi call: 2015-11-23
It is really important for BCO to allow a specimen to be the input of a specimen collection process for sub-sampling. Suppose you collect one specimen, then take an aliquot of it, which also results in a specimen. I think it would be confusing to call those different processes, although the latter could certainly be a more specific subclass of the former.
'material portioning' (alt label: 'aliquoting') would be the process of
splitting something into multiple smaller portions. And that can be done
with specimens, but wouldn't be considered the 'collection' process.
On Tue, Nov 24, 2015 at 8:33 AM, Ramona Walls rwalls@users.sf.net wrote:
--
Bjoern Peters
Associate Professor
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
Related
OBI Terms:
#778If material portioning of different kinds (aliquoting, slicing tissue) is the only way specimen families are created, then based on Bjoern's suggestion, I'd say we can make 'specimen family creation' a subclass to 'material portioning'.
If this is not the case, then I'd suggest to tread 'specimen family creation' as an attributive collection of processes of a different type.
I think specimen families can only be generated through 'material processing'. The 'material portioning' is one kind of material process.
Here is my proposal: change the definition of 'specimen family creation' to:
A planned process that has as its input a specimen and as its output a collection of specimens through one or more material processing process(s).
Propose to add logical axiom:
has part some 'material processing'
It is still under discussion whether has_part a process is equvilent to subClassOf a process. The question raised in the assay harmonization. We can discuss about it in the call when Bjoern can join.
So, based on today's OBI meeting, the proposal was:
'specimen family creation'
Superclass: 'material processing"
Definition: A material processing that has as its input a specimen and as its output a collection of specimens.
Logical axiom: 'material processing' and has_specified_input some specimen and has_specified_output some specimen set
Example of usage: Aliquoting one specimen to multiple tubes, slicing a tissue specimen into multiple sections.
Discussed on Feb. 2, 2016 call. No objections to adding but two points were raised. Will add this and close the tracker item.
1. In addition to textual examples, it is helpful for logical axioms to provide an example on how it will be used (i.e. with RDF). The major use case (in my view) is to be able to define and relate specimens in a family to their source.
2. Specimen family creation as discussed at the beginning of this thread may involve multiple material processings.
Terms discussed and agreed upon have been added.