Moore Notes 9 18 13

Group Call

• Participants: Katie, Jonathan, Dongying, Guillaume, Sarah, Ladan, Josh, Stephen, Tom

• Morgan Langille got a faculty job at Dalhousie

• Dongying's marker paper will be published in PLoS ONE

• Josh will talk to Alexis in next week about Tara oceans sampling

• Tom talked to Morgan about PICRUSTs
  • We would need to thoroughly vet its application to marine communities
  • Good idea to compare shotgun metagenome protein abundances vs. PICRUSTs estimates
  • Limited to taxa in the GreenGenes Db
  • Limited to proteins families in KEGG (or you need to annotate genomes with your protein db)
  • Doesn't do any correction for genome length, unannotated proteins, etc.
  • Jonathan: Marine is probably the next best place to go due to GBMF funded genome sequencing
    • see https://moore.jcvi.org/moore/
    • also see http://camera.calit2.net/microgenome/
  • Guillaume & Tom want to start with English Channel L4 dataset
    • Need to annotate with KEGG (did SFams already)
    • Guillaume is compiling the 16S from same samples and running it through PICRUSTs
  • Caveat: 454 sequencing, so lower coverage than enteric samples in PICRUSTs paper

• Stephen has been exploring normalization of protein family abundance measurements
  • Adjust for number of genomes in the community
    • To account for coverage of proteins in community by db
    • Also to adjust for genome size
  • Ground truth with Dongying's marker genes
  • Caveat: doesn't account for archaea or eukaryotes very well right now, but could in future
    • Tree of Life project has eukaryotic (protist) marker genes - universal, but not necessarily single copy
    • Still won't know average genome size very accurately