Moore Notes 3 9 10

Group Call

• Steve: Did we track down the GOS PCR 16S data (Shaw et al)?

• Steve: Is ZORRO available for masking Guillaume's alignments?

• Every one should check that they are correctly signed up for the iSEEM Google Group

• Dongying research update: http://bobcat.genomecenter.ucdavis.edu/dwu/DW0309.pdf
  • identifying marker genes for specific phyla
  • can combine (bottom up) to expand AMPHORA set
  • question: how to set a threshold for hits to profile HMM in a particular clade (below each node in the tree) - time consuming too
  • Martin: there is no totally reliable way to do automated identification of subfamilies
  • Tom: there are automated methods (e.g. PSYPHY) to determine cuts
  • Also working to develop a score for phylogenetic informativeness of a marker (does it produce monophyletic groups?)
    • this is essentially comparing the gene tree to the genome tree (e.g. AMPHORA)
    • other ideas:
      • entropy,
      • in exponent divide by maximum possible distance = sqrt(n*(n-1)) rather than n,
      • parsimony (e.g. number of LGT events needed to explain tree),
      • scale by number of taxa in tree

• JG: update on Kelly's visit to U of O