Semantic resources project/Use Cases/Meeting Notes/SWAN/Meeting 10072009

Meeting with Elizabeth and Gwen 10/7/2009 -- 10am. In attendance: TWD, EW, GW

AlzForum started in ‘96
 * Basic HTML, started posting papers, building an online community (with “June”)
 * Beginning in 2002-2003, started thinking about organizing information not by “form,” but by “function” – papers, not genes, etc.
 * SWAN was their “knowledge-management” project
 * SCF was rebuilding AlzForum in CMS style – first StemBook, then PDOnline, now AlzForum again.
 * Porting:
 * Research Model page
 * Antibody page.

They want to talk about the “Compendia” section – “the CNN for Alzheimer’s research”.
 * Editors: June, Gabriele, Tom Fagan
 * Scientists write “comments” actively, but these get edited too.
 * RSS feeds for specific pages.
 * Profile pages for researchers

Compendia:
 * Antibodies
 * Central hypothesis focuses around plaques
 * A-beta plaques
 * But is it the plaques, is it the soluble form, small dimers and trimers, etc.
 * APP-APLP antibody pages – an example of well-curated antibody database
 * Where should the joins occur?
 * Paper references
 * Reactivity column: papers – reactivity and antibody references are important separately.
 * Papers:
 * Links to pubmed
 * Links to primary news, related news
 * Researcher comments!
 * powID – “papers of the week”
 * “pmID=?” also works in the URL”
 * some papers aren’t in pubmed but are powID’ed.
 * Protocol database
 * Linked to antibodies
 * Linked to PDFs
 * Lab, PI, submission, METHOD
 * Mutations
 * APP, Tau, Presenilin Mutations
 * This database doesn’t change as quickly
 * Schematic and Mutations table
 * One form of the table is on AlzForum – another is on a group from van Broeckhoven lab.: AD&FTD database.
 * AlzForum focuses on mis-sense mutations
 * AlzForum has an extra column – “age of onset”
 * Research Models
 * Table of transgenic mice
 * All mouse, no drosophila or worm
 * Protein isoforms – APP slides (ask for these!)
 * Mouse TG 2576: freely distributed, thousands of articles
 * Human cDNA with two mutations (including “the Swedish Mutation” which ramps up beta-cleavage)
 * Karen Hsiao, Science, 1996 (and an earlier one in Neuron in ’95)
 * But the earlier strain was SJL, and not good…
 * Taconic C57BL6 is very different from the Jax lab version
 * Mouse genealogy trees -- ?
 * Apparently the strains genealogy is very important.
 * The chart doesn’t appear to be important.

GLUE code:
 * update NC from SWAN maybe once a month.

Publications as hubs in the connective network
 * Link to PDOnline?

Representation issues:
 * Representation of identical claims
 * STRAWMAN tag?
 * “developed_from” relationship between hypotheses?

Pathogenic Narrative:
 * Initial Condition
 * Perturbation
 * Pathogenic Event
 * Pathologic Change
 * (Post-therapy stage?)

Evidence Type Ontology:
 * Genetics
 * Epidemiology
 * Mechanism/Structural Biology
 * Physiological Systems
 * Biomarkers
 * Target Validation
 * Animal Models
 * Clinical Trials
 * Etc…

Evidence Maps
 * every claim gets a narrative and an evidence type annotation, and their evidence map then cross-cuts the claims and hypotheses against these categories.

“Research Questions” – linked to the type of experiment.
 * Questions asked by users