Notes 20100629.120000 CS 798 Andre's Talk

From SnOwy - Ed's Wiki Notebook

Contents 1 Horizontal Gene Transfer 2 Background 2.1 The present work 2.2 Aside

Horizontal Gene Transfer

• because of a horizontal transfer of genomic material, a phylogenetic reconstruction will incorrectly map a particular taxon to the wrong branch of the tree
• mechanisms
  • "eating" strange DNA found scattered in the aqueous solution
  • conjugation
  • incorrect virus genome clipping

Background

• want to describe the genes that are inherited vertically
• want to distinguish but correctly reconstruct history of horizontally transferred genes
• orthology problem

The present work

• bimodal distribution of BLAST scores when the model E. coli were used vs. Salmonella, vs. Vibrio
• a mode exists in the 90% range of the blast scores, where 12% of the genes exist (for E. coli vs. Salmonella)
• a mode exists in the 10% range of the blast scores, where 6% of the genes exist (for E. coli vs. Vibrio)
• the stuff that occurs below 30% identity are the "non-specific activity" matches -- things like kinases, phosphorylases etc.,
• several topologies produced, after throwing out all but 205 domains (these domains may not have more than one duplicate in the genome)
• rRNA is thought not to transfer horizontally, that's thrown out too
• phylogonies produced end up raising questions-- there is a biotin pathway gene that is transferred but it's not clear why the transfer was more successful than the existing biotin gene

Aside

• Able to separate species of bacteria based on horizontal gene transfer rate -- either due to geographical distance or due to some mechanical limit (Great Dane / Chihuahua).