Needs Analysis Workshop/Breakout group 1

From phenoscape
Revision as of 17:48, 17 September 2007 by Jpb15 (talk | contribs)

Break-out Group 1: Developmental genetics of morphology

Examples for driving research questions:

  • Which genes are known to be expressed in the development of a particular morphological structure?
  • Is there a model organism mutant that has a phenotype similar to a human disease?

Characters

GW - describe natural units - not necessarily arbitrary - can phenotype be decomposed into natural mechanistic units

CK - could ontology hold you back from describing some things

GW - can lead to confusion

EJ - this database can be a way of discovering these units

GW - would database be like GenBank - I could describe a fish and deposit my phenotypes - structure of data would be fluid

PM - data is representation of current knowledge - can change over time

GW - how is change in knowledge, like of orthology statements, handled in genome databases

Manually updated, typically

MW - We should move above mechanics and focus on biological questions

EJ - how can we use data to identify discrete modules of variation within evolution

GW - taxonomic work in mammals often involves tooth characters - usually cusps are individualized - actually character identity should be attached to tooth type rather than cusp as unit character - resolves many homology issues

Question: See same tooth type evolving again and again. Developmentally, cusps don't come and go individually. - see covariance within mutants query for mutants that affect cusp X - always see that other cusp is affected


GW - Homology and analogy - often too "yes or no" - instead could have "character model" - contain kind of cells within, location in body, relation to other characters, what kind of states can it assume - represent variational tendencies of a chunk of the body - avoid simplifying biological reality - open-ended structure