Tara Oceans Analysis Plans

Tara Oceans Analysis Ideas

• Prevalence of novel SFams (diverse families with no annotated function) across Tara sampling locations.
  • Potentially correlate with annotated families and/or environmental data, location, etc.
  • Team: Aram, Stephen, Stacia
  • Approach: Pipeline of scripts using diamond and novel SFams db written by Stacia
  • Status: Complete - Stacia is writing up manuscript

• Global niche modeling of distributions of specific gene families and pathways.
  • Genes/functions of interest for focused analyses:
    • Antibiotic resistance and synthesis
    • CRISPRs and related proteins
    • Photobiology
      • Photosynthesis
      • Light receptors
      • UV DNA damage protection and repair (Eisen)
      • Proteorhodopsins (Lizzy Wilbanks, Sarah Hird)
      • Circadian rhythms (Eisen)
    • Carbon fixation
    • Iron scavenging
    • Nitrogen cycle - nitrogen fixation, nitrification/denitrification, ammonification, anammox (Adrienne)
    • Biosynthetic pathways of biomedical relevance (e.g., http://elifesciences.org/content/4/e05048)
    • Petroleum Hydrocarbon and Plastics degradation proteins
  • Starch utilization system operons (selfish bacteria; Carol Arnosti) - do the genes in the operon co-occur? Do other genes co-occur with them?
  • Potentially explore historical environmental data versus current to look for lags similar to what we see in soil and to predict extinctions.
  • Possibly look at overall diversity metrics to predict hotspots.
  • Team: Stephen, Josh, Patrick?, Adrienne, Carol
  • Approach: ShotMAP with KEGG db, model selection, predictions/maps, interpretation
  • Status: ShotMAP done, rest to do

• Strain-level analysis (copy number variants, single nucleotide variants) of prevalent species across Tara sampling locations.
  • Build phylogenies.
  • Potentially look at gene prevalence within species and correlate with environmental data, location, etc.
  • Team: Stephen, ?
  • Approach: PhyloCNV
  • Status: Started - Stephen ran PhyloCNV but probably needs help assimilating and interpreting results

• Ecological annotation of protein families
  • How do proteins interact with the environment? Use KEGGs and SFAMs from above annotation frameworks and correlate each family with environmental data
  • Is ecological covariation a predictor of pathway interaction? Quantify correlation between KEGGs that linked through pathways across sites.
  • These ideas are related to several listed above and integrated analyses may be more efficient
  • Team: Tom
  • Status: To do