Moore Notes 1 21 15

Discussion of TARA Oceans data

• Participants: Katie, Josh, Tom, Guillaume, Stephen

• Stephen:
  • Data embargo issue
  • Updated summary (slides)

• Analysis discussion:
  • What do we want to do with the data?
    • Start with aims of proposal
  • How to preprocess?
    • They will likely release EGGNOG abundances
    • They may map reads to assemblies (gene catalog)
    • Do we need something more/different?
      • Database
      • Classification thresholds
      • AGS normalization
  • diamond vs. rapsearch2
    • Do a quick comparison (correlation) of bit scores
      • If highly correlated, can use previously identified thresholds
  • Many (667) samples to process
    • Prioritize the prokaryote size fraction, then protists, then viruses
    • Prioritize open ocean (all?), surface waters (approximately 216 samples)
    • Start with metagenomes
  • Josh will look at ecological variability (MESS plots)
    • Can we do global predictions?
    • Are there samples we would drop and therefore do not need to run for read classification?
  • How much QC is needed
    • Stephen: Probably hasn't been done, but also not necessary
    • Better to keep track of quality and use that info downstream
    • Illumina looks better than 454
    • Could QC one library and compare protein family abundances pre and post QC
  • Size fractions reliable?
  • Stephen will start AGS analyses right away