Moore Notes 2 4 09

Group Call (KP and JG away)

• Here are some examples of the alignment and trees for the DeLong SSU rRNA, built with a variety of methods.
• At this point we're hashing out mechanics and squinting to see if the sequences are reasonable.

(NJ (FastTree), parsimony (PAUP), likelihood (GARLI), Bayesian (MrBayes)).

• • File:DelongSSUali trees.zip

Srijak emailed nucleotide sequences out for Delong marker genes.

• Nucleotide sequences are aligned, trimmed and masked.
• There are redundant sequences in nucleotide and amino acid datasets.
• Redundant sequences have different identifiers.
• For now, Srijak will email a list of the redundancies.

Simulations:

• Should we use peptide or nucleotide data to test building approaches with simulated data?

• What genes should we initially run simulations on?
  • recA
  • RPOB

• What methods should we use to compare trees?
  • Robinson-Foulds
  • Quartets
  • Nodal: [1]

• Sam will try to get initial marker gene simulation dataset be early next week.
  • Wants to use a set of genes from one gene family that are reasonably diverse.
    • • Could use:
        • Amphora sequences
        • Dataset from Martin
        • Get script from Dongying that will select sequences that maximize diversity from a dataset.

• Jess B will concurrently be working on 16s rDNA simulations.

Another Skype call tomorrow at 1.