Moore Notes 3 16 15

From OpenWetWare
Jump to navigationJump to search

Group Call

  • Participants: Katie, Tom, Patrick, Stephen, Ladan, Josh, Sarah
  • Tara Oceans analysis:
    • Running samples (Stephen)
      • Stephen found a bug in rapsearch (authors looking into it)
      • Tom will conduct a run with old version and shotmap batching
    • Filter sizes (Josh)
      • Five different sizes (overlapping)
      • Most sizes not done at most geographic locations (example)
      • Size 0.22 to 3.0 is most common, 0.22 to 1.6 used in other locations
      • Confounded with location, so cannot model out
      • Stephen: filter size is positively correlated with average genome size
      • Josh will spot check for any missing files and summarize
    • Powers et al. (2015) natural products paper
      • Used eSNaPD:
      • Patrick: coverage may be lower in non-targeted Tara Oceans data
      • Tom: disentangling taxonomic signal would be a good idea
      • Tom: look at hits on a gene family phylogeny
      • Josh: places with highest level (>1% of reads) of each gene family cluster
        • But should not be called "hotspots" since only at sample locations and not global
        • Survey gap analysis: predicting where to sample to find different functions from previously known
      • Which protein families and marine systems to focus on?
        • Quorum sensing (too diverse?)
        • KS domains good because same domain used for many different functions
        • Non-ribosomal peptide synthesis
        • Peter Dornstein: mass spec database (BLAST for peak profiles) to prioritize
  • Next calls
    • March 30: Patrick
    • April 13: Tara Oceans and Sharpton lab clade diversity project