Goal

• model validation: compare the priming experiment HR-SIP sequencing data to simulations of that same data

Method

• Find taxa that are closely related to sequenced genomes
  • relatedness based on 16S rRNA
  • CD-HIT: OTU rep sequences & complete genome 16S seqeuences
    • cutoff = 97% seqID
  • Select a particular 13C gradient (& matching 12C control):
    • Infer total abundance of each target taxon
    • Infer total richness of starting community
    • Get distribution of total OTU abundances per fraction
• HR-SIP simulation with target taxa abundances set by Priming Exp
  • Fill in rest of community with pseudo-genomes
    • pseudo-genomes are just fragment-length/GC distributions
      • GC distribution as a normal
        • GC_dist_mean is drawn from normal distribution (mean=50, sd=7)
    • number of pseudo-genomes: gradient community
  • abundance distribution: lognormal
  • total abundance: 1e9
  • subsampled: based on Ashley's dataset

OLD

• Use the genomes of target taxa to simulate the subset of the HR-SIP dataset
  • Use inferred
• Fill in rest of community with pseudo-genomes
  • pseudo-genomes are just fragment-length/GC distributions
    • GC distribution as a normal
      • GC_dist_mean is drawn from normal distribution (mean=50, sd=7)
  • number of pseudo-genomes: 5000?
• Simulate HR-SIP data
  • abundance distribution: lognormal
  • total abundance: 1e9
  • subsampled: 20k?