See also: IRC log
<vipul> set logs, world-visible
<vipul> Helen's page for Parkinson's Disease http://esw.w3.org/topic/ParkinsonsDisease
<vipul> http://esw.w3.org/topic/ParkinsonsDisease
Vipul: intro - context - Don, Helen, & Vipul - discussed on ACPP TCon to extend current UC to bedside (see Helen's PD URL)
DD: see PD UC - biological causes
w/cites - discussed associtated dimentia & olfactory
impairment - many specific to PD
... locus coeruleus damage - correlate neuron loss w/NE
depletion & dimentia - there is a lot of subtlety to
interpreting phenotype correlations
... e.g. - S.nigra loss associated with motor problems - e.g.
dyskinesia
... l.c. loss appears related to cognitive impairment - memory,
executive functions, visual-spatial mapping, language,
etc.
... l.c. AND nuc. basalis of Meynart - associated w/biochem
deficits & cog. impairment
... still roughed out - but will continue to enhance
VK: can you step through from molecular effect to dimentia - covering intervening entities?
DD: Oooh - tough - try executive
function deficits - problems (e.g., like following company
protocols or manners - inappropriate outburst) - this
correlates with decr. Choline Acetyle Transferase (ChAT) in
frontal crtx - therefore neurochem. Rx for low ChAT may assist
in recovering exec. func.
... remember any specific Use Case (UC) glosses over many
related issues - must limit UC
VK: glossing fine, as long as we can provide additional value through links derived from those entities that are covered in the UC
Amit: what is the role of literature informatics in this semantic approach to describing the data related to an UC
VK: our ontology is spare - mostly basic bio science - not clinical & medical, yet. HC working on this - focus of demo is to use avaialble primary and accessory data repositories - data FACTS - as opposed to their distillation in the lit.
Amit: tools to create idea graphs by extracting UMLS & MeSH terms from a defined lit. corpus; also - if you have multiple data sets in RDF, can do correlations across graphs - extension to SPARQL to specify constraints based on specified sub-graphs and node traversal limits.
<helen> three well established nuroscience literature corpus:brainmap.org
<helen> three well established nuroscience literature corpus:brainmap.org
<vipul> Thanks, Helen!
<helen> trying to get the other two sites
VK & Amit: have a focussed working session on UC, specific queries & data sets (early next year) - include students from Amit's lab to build demo - based on F2F UC w/known answers as a metric
VK: well-known migraine study by
Swanson - can use VK lab tool to automatically derive
info
... Amit tool semantic mining - we are focussed on semantically
well-founded data integration now.
Amit: see those 2 as intimately linked
VK: agreed - but focussing first on integration
HC: focus on PD clinical
guidelines - have references to specific dtb ref numbers - use
those to assemble biological causes
... symptoms = clinical features in the clinical
guidelines
... drew entity graphs for these features
VK: how link to what DD is doing
HC: Rx side - different types of
Rx - medication in particular symptom (did already for
depression) - could pick PD dementia - look at pharm Rx - link
back to biochem. based on known biological role of pharm
agent.
... each feature comes with a model - hypothesis including
where & how of biological entities involved.
DD: you fleshed out PD depression & some of the olfactory symptoms - could you do this with dimentia
HC: need to tie in neural systems pathways
DD: pathways = connections from one brain region to another
BAMS: http://brancusi.usc.edu/bkms/
BAMS XML dumps: http://brancusi.usc.edu/bkms/bamsxml.html
HC: tie behavior to molecules - e.g., olfactory oderant links to olf-receptor - then on to biochem, specific neurons in the Olf Epithelium and connectivity (see SenseLab for support in making this leap)
DD: put out more detail in the biological causes
HC: work with DD to add more
links from clinical features to dtb refs - represented as
"triples"
... go to literature to add some supporting info for diagnosis
- weight assertions based on certainty of observations found in
lit.
VK: when to meet next?
HC & DD: will meet next week to reconnoiter - determine whether covered all the available details
CocoMac - Untangling the Brain - http://134.95.56.239/home.asp
DD & HC: work more on the BMC Bioinformatics manuscript
This is scribe.perl Revision: 1.127 of Date: 2005/08/16 15:12:03 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) No ScribeNick specified. Guessing ScribeNick: BillBug Inferring Scribes: BillBug WARNING: No "Topic:" lines found. Default Present: Vipul_Kashyap, Bill_Bug, John_Barkley, Don_Doherty, Helen_Chen, Davide_Zaccagnini, Amit_John Present: Vipul_Kashyap Bill_Bug John_Barkley Don_Doherty Helen_Chen Davide_Zaccagnini Amit_John WARNING: No meeting title found! You should specify the meeting title like this: <dbooth> Meeting: Weekly Baking Club Meeting WARNING: No meeting chair found! You should specify the meeting chair like this: <dbooth> Chair: dbooth Got date from IRC log name: 12 Dec 2006 Guessing minutes URL: http://www.w3.org/2006/12/12-hcls-minutes.html People with action items: WARNING: Input appears to use implicit continuation lines. You may need the "-implicitContinuations" option. WARNING: No "Topic: ..." lines found! Resulting HTML may have an empty (invalid) <ol>...</ol>. Explanation: "Topic: ..." lines are used to indicate the start of new discussion topics or agenda items, such as: <dbooth> Topic: Review of Amy's report[End of scribe.perl diagnostic output]