Fwd: [Fwd: DENRIE branch review]

[Chris Stoeckert]

Branch review from James. (Thanks James!) We also had one from Bjoern to discuss. Will add them to our mail list once I remember how - let me know if any of you have gone ahead and beat me to it.

Thanks,

Chris

Begin forwarded message:

From: James Malone <mal...@ebi.ac.uk>

Date: March 6, 2008 6:57:59 AM EST

To: Alan Ruttenberg <alanrut...@gmail.com>, Chris Stoeckert <stoe...@pcbi.upenn.edu>

Subject: [Fwd: DENRIE branch review]

Hi Alan, Chris,

I tried to submit my branch review to the google list but I'm not a member (I've just requested membership now).  Could you possibly forward this on from me please (below)?  Thanks.

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

From: James Malone <mal...@ebi.ac.uk>

Date: March 6, 2008 6:55:21 AM EST

To: obi-denr...@googlegroups.com

Cc: mal...@ebi.ac.uk

Subject: DENRIE branch review

Dear DENRIEers,

As agreed on last weeks dev call, here is my review of the current DENRIE branch.  I was pleasantly surprised by the amount of progress made, it's been a while since I've really looked at the branch.  I should try and say in an unpatronizing way as possible, great work guys :)

Specific comments (please feel free to bounce queries/comments to me):

-Missing curation status, I think almost all classes from "non-realizable information entity" down.

-Also, definition_source is missing from most classes.  This is also part of minimum metadata.
-Class "source code module" needs to be linked to algorithm presumably (when algorithm is sorted ;) )

-Moreover, referring to the above  "source code module" class, I was at first confused why this is not a sibling or child of "text based digital entity".  An example of the former is given as 'Perl Script' and the latter ' C++ Source code file'.  It could just be my misinterpretation of what the "source code module" is trying to describe.  So my guess is that what you are saying by this is that "source code module" class is an encoding of an algorithm in programming language and this can then be encoded_as a "text based digital entity" for example, or simply written out on paper or on a chalkboard.  Is that about right?  It may just be the examples that have thrown me because I can understand the latter reasoning.  I presume a relation such as is_encoded_as is to be added or similar?

-It took me a little while to get my head around the idea of a "narrative object" being a set of propositions and a graph being a type of narrative object.  I can see the reasoning but perhaps the definitions for narrative object could be loosened slightly to ensure it is intuitive that graphical objects fit hear too.  I think it is the idea of a 'set of propositions' that fooled me.  Perhaps if it contained the word 'describes' somewhere this could help, e.g. "A narrative object is an information content entity that describes a set of propositions.".  More confusing for me was that we have class "image" as a sibling of "narrative object".  Is a graph an image?  By the definition of image ('provides a visual representation of for the result of a measurement.') I would think it is.

-Question: we have "measurement unit" in as a class, are we actually covering that?  Will this be an import from somewhere else eventually?

We need to link to some of these classes from the Data Transformation branch, I should probably make a list and submit what we need I guess!  I'm happy to help with editing/adding these though and help on the branch.  Hope this review is useful in some way anyway!

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