Genome size and Complexity

Genome size and Complexity
(as told by Michael Lynch)
Genome size and complexity varies
across the tree of life
Lynch 2007
Some Big Questions
• What is the relationship between genomic
and organismal size/complexity?
• Are genome size changes adaptive, or
passively acquired?
• How do these changes occur
• How do study all this?
Proximal mechanisms of genomic
Mobile element proliferation
Segmental duplications
Strand slippage
DS DNA breaks
– Insertion propagation (biased gene conversion)
– Microdeletions (nonhomologous end repair)
• Unequal crossing over
• Illegitimate recombination
• Selection for or against such modifications
Explaining variation in genome size
• Adaptive, neutral, deleterious
Selfish DNA hypothesis
Bulk DNA hypothesis
Metabolic cost of DNA
Petrov neutral hypothesis
Mutation Hazard Hypothesis
Lynch and Conery 2003:
“Mutation Hazard” Hypothesis
• Body size / complexity increase
• Reduced Ne
• Reduced efficacy of selection
• Non-coding DNA proliferates
• duplicate genes retained
Increases in genome size and complexity are a drift-driven
consequence of reducing efficacy of selection
Neu scales inversely with genome size
Lynch and Conery 2003
Half life of duplicate genes greater in larger genomes
Lynch and Conery 2003
Intron number and size greater in
larger genomes
Threshold effect
Lynch and Conery 2003
TE size/number scales with
genome size; threshold effect
Whitney and Garland, 2010
Whitney and Garland, 2010
Whitney and Garland, 2010
Conclusion: no mechanistic connection between Ne and genome complexity
Subtext: Adaptive processes account for variations in genome size?
Lynch 2011 rebuttal
• Neu has no shared evolutionary history
– Comparative methods inappropriate
• Topology and branch lengths of phylogeny
is suspect
• Threshold effect is most important
prediction of MH; not addressed in W&G
• OLS is unbiased
Whitney et al. counterpoints
• Significant phylogenetic signal (K) for all
traits examined (genome size, etc)
• Life history and other factors underlying
Neu can and do have phylogenetic signal
• Consistent results with different
• Too few data to conduct threshold tests
• OLS clearly biased

similar documents