What we shall have to find out is how we use Hyper-G to construct glossaries.
In PPS95 I wrote series of templates via FORMS (for term, index, etc.)
There must be something similar (and probably better) in Hyper-G. Note
that there is a powerful search facility (try , say, phenylalanine). The
main drawback is that the display of 300 terms in a vertical line at the
same level is messy, so clearly we need some concept analysis. (What
this means is the grouping of similar terms into a subtree - e.g.
aminoacids, proteins etc.) This could be a useful thing for the new
curators to think about :-)
I have updated the home pages and there is quite a lot that is new - not
just rewritten.
------
I have downloaded the MARTIF material from its home page in Brigham Young
Univ. and extracted the examples and the DTDs. (Gerhard - these were in
*.zip and appeared to have some DOS stuff like ^M, \376 (thorn?), and
some missing whitespace (? tab problems). However I got it to run under
sgmls and display under my costwish browser where it looks quite nice. I
can't get the picture to you for tomorrow - hope all goes well.)
In the medium term I think we eant to create the following:
- A simple template for generic glossaries (like PPS). (NOTE, pictures,
molecules, etc are not IN the term, but pointed to from it.)
- A means whereby the template can be embedded in Hyper-G so that people
enter terms without worrying about syntax. Alternatively the term can be
submitted in toto through FORMS (but NOT edited that way).
- A means whereby curators can edit entries still using the template
- A means of posting annotations to entries.
(I have no idea whether Hyper-G supports the authoring of SGML - we'll
have to find out.)
------
Here are few specific querstions.
(1) How do we enter case-sensitive terms (e.g. PCP and pCp?). I am not
clear what MARTIF suggests
(2) How do we deal with truncation? Up to now we (crudely) input appropriate
parts of speech (e.g. leucine, leucines, leucyl, etc.) In HG you can
search for leuc*. Do we still need to add all these terms or can we
enter leuc* as an indexing term??
(3) I would like to resurrect the ideas from Collaborative Clickable Biology
so that we have clickable concepts. In CCB I prototyped this with a
genome (HIV) and a metabolic cycle, but it could apply to anything.
We'll have to see whether Hyper-G allows users to draw hotspots on a
remote imagemap and add URLs to them (I have a vague recollection it
might allow that...)
(4) What do we use for uniqids? I am keen that we use the most obvious
term, all in lower case and with underscores for whitespace. Only
alphanumerics should be allowed:
ramachandran_plot
criminal_justice_act
virtual_hyperglossary
interleukin_1_alpha
and so on. Of course the user can also search for other indexing terms
so this is just to help.
I'm looking forward to hearing from you...
P.
Peter Murray-Rust, Glaxo Research & Dev. (pmr1716@ggr.co.uk); (BioMOO: PeterMR)
Birkbeck College, ubcg09q@cryst.bbk.ac.uk, CBMT/Daresbury mbglx@seqnet.dl.ac.uk
http://www.cryst.bbk.ac.uk/PPS/index.html, http://www.dl.ac.uk/CBMT/HOME.html