Moore Notes 7 30 14

From OpenWetWare
Jump to navigationJump to search

Group Call

  • Participants: Katie, Tom, Dongying, Guillaume, Sarah, Stacia, Stephen, Josh, Patrick
  • MMI event at San Jose tonight
  • Shotmap functional analysis
    • Analyzed L4 and MetaHIT
    • Results: slides
    • Functional diversity vs. phenotypes
      • Kegg KO Shannon entropy low in Crohn's disease (CD) metagenomes
        • 3-group tests (CD, UC, healthy) not significant, but CD significantly lower than healthy
      • SFams richness and Shannon entropy low in CD
        • How accurately can we estimate richness?
        • Diversity may be correlated with BMI because both are low in CD
    • Principal components analysis
      • Center and scale or not?
      • Families with high loadings on PCs (ECF factors, iron membrane receptor)
    • Testing for significant differences in abundance between groups
      • Families with lowest q-values
      • Check if family names have error
    • Need to compare across databases
    • Comparison to taxonomic profiles
    • Differentiation from MG-RAST, humann, etc.
      • M5NR is pretty diverse
      • But shotmap allows a custom db (should show this, e.g., Morgan et al findings, extended metaCyc)
    • Link to simulations
    • Review literature on MetaHIT re-analyses
    • Strength is a tool that works and simulations to evaluate it and optimize
  • Upcoming calls:
    • Aug 6: Stephen
    • Aug 13: Patrick - biomarkers
    • Aug 20: Josh/Tom - allergens
    • Aug 27: Dongying - phyeco