Example: Ortholog Prediction

1
Example Ortholog Prediction

• Orthologs two genes related by speciation events
  alone. the same gene in two species,
  typically, same function.
• Paralogs two genes related by at least one
  gene-duplication divergence event.
• Homology an ortholog or a paralog?
• Experimentally very hard to answer.

2
Gene Evolution through Speciation vs. Duplication
speciation
gene duplication
orthologs
paralogs
3
Using Multiple Genomes for Ortholog
Cross-Validation

• Within a given genome, ortholog should be more
  similar than paralog (same time of divergence,
  but divergent functional pressure on paralog).
• Complete genome if theres an ortholog, youll
  find it!
• Multiple genomes quickly sort out
  lack-of-ortholog, multiple paralogs, gene-loss
  problems that could lead you astray.

4
Cluster of Orthologous Genes
Orthologs should be reciprocal best hits.
?
Tatusov, Koonin Lipman, Science 278, 631 (1997)
5
Reciprocal Best Hits Indicate Orthologs
Reciprocal best hit
Non-reciprocal best hit
6
Multiple Genomes Screen Out Errors
True orthologs should give consistent,
reciprocal best-hit pattern as more
genomes added.
?
Chance of missing a true ortholog, or predicting
an incorrect ortholog, exponentially decrease
as more genomes added.
7
Simple Ortholog Cluster (found in three genomes)
KatG
sII1987
YKR066c
Catalase peroxidase in E. coli, yeast,
Synechocystis, from Tatusov, Koonin Lipman,
Science 278, 631 (1997)