ISO 20691 draft - Please comment and revise
Golebiewski, Martin
Dear all,
As we just have discussed the ISO 20691 draft in the Annotation session at COMBINE 2020:
“Biotechnology — Requirements for data formatting and
description in the life sciences for downstream data processing
and integration workflows”
Please review the document and give me your feedback asap. The document is supposed to be sent to the ISO central secretariat at the end of the year to start the publication process. If you like to have an editable version, please let me know.
The draft standard also refers to the COMBINE qualifiers in clause (“Semantic criteria and requirements”) and to “Requirements for ontologies suitable for annotation of biological data” (clause 7). Please also review the community formats and standards referenced in Annexes A and B.
Thanks and best regards,
Martin
------------------------------------------
Martin Golebiewski
HITS gGmbH
Schloss-Wolfsbrunnenweg 35
69118 Heidelberg
Germany
phone +49 6221 533 281
fax +49 6221 533 298
email martin.go...@h-its.org
_________________________________________________
Amtsgericht Mannheim / HRB 337446
Managing Director: Dr. Gesa Schönberger
Scientific Director: PD Wolfgang Müller
John Gennari
Thanks much, Martin. This document is helpful.
One of the (several) outcomes for me from COMBINE 2020 was to try
to update and improve the understandability and usability of the
biomodels.net qualifiers. (See
http://co.mbine.org/specifications/qualifiers)
The document you provided is relevant, but seems to be a bit
different in scope. Namely, the focus of your ISO document is on data,
whereas the biomodels qualifiers are more specifically about
annotating a biomodel.
The relevant part of your document, as I understand it, is Table
2 (beginning on p. 25). However, I note that this table is a list
of "examples of predicates", rather than any sort of prescriptive
requirement or minimal list of required predicates.
Elsewhere, the document talks about how these predicates "should
be used" for data annotations. Of course, Table 2 isn't quite the
same as the list of biomodels qualifiers... But this makes sense
if the focus is on the data, rather than models.
Thus, I suppose that we should just look at this document as
another source of ideas for how we might scope or limit the set of
qualifiers.
Let me know your comments and if I've misunderstood anything. It
was great to "see" everyone at COMBINE.
-John Gennari.
--
You received this message because you are subscribed to the Google Groups "combine-annot" group.
To unsubscribe from this group and stop receiving emails from it, send an email to combine-anno...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/combine-annot/86E9833C-C473-4AA2-90F9-53AA07001BEF%40h-its.org.
Golebiewski, Martin
Dear John,
Dear all,
The standard clearly states in its scope that it refers to all data in the life sciences, including models as a specific type of “data”, without mentioning it explicitly. To make it more clear, that it does not distinguish between data and models, we will include an explicit hint in the scope that the term “data” includes also computer models.
There are several parts in the document draft that are highly relevant for models and especially for their annotation. In particular, all clause 6 (“Semantic criteria and requirements”) regulates the annotation of data (including the annotation of models) in a real normative way. It actually do provide a minimal list of required predicates for the annotation (table2) and it would be great if you could check and suggest additional ones if required. It also provides a minimal list of required biological descriptors (Table 1: “Examples of basic required biological descriptors”). The wording “examples” is an ISO-specific way of standardizing things, without making it mandatory. Moreover, “Requirements for ontologies suitable for annotation of biological data” (clause 7) provides some high-level recommendations for biological ontologies, also used for the annotation of models. So, please check carefully, if all requirements from modelling side for ontologies/controlled vocabularies are addressed. If anything would be missing or the text should be modified or extended, please edit using the commenting table attached (a filled example is also attached to demonstrate, how it should be filled).
The annexes A and B of the standard provide comprehensive guides for formatting and annotating data/models and refer to the most important community standards. For modelling these parts are extremely important (but others might be too):
A.2 Formats for computer models of biological systems
A.3 Formats for model simulations and their results
A.4 Descriptors for quality measurements for data and models
And all annex B (“Minimum reporting standards for data, models and metadata”)
So, you all should check, if all important standards for modelling and annotation of models are covered there and are described in a comprehensive way. If anything would be missing or the descriptions should be modified or extended, please edit these, using the commenting table attached.
I would like to stress that it not provides just a source of ideas, but will be binding as standard. Your reading and comments, however, demonstrates that we might change the wording and also might consider extending the content to clarify both, the scope and the rules. Your input and feedback, and the input of all in the COMBINE-annot mailing list might help to improve the draft and make it better understandable, also for modellers and/or tool developers that are part of the target group of this standard. So, I would like to invite you to edit the document and propose changes and extend the content, where needed.
Hope that clarifies it a bit. Thanks for your input and feedback.
Best regards,