Genome evolution

• There are both proximate and ultimate explanations in molecular biology

• Mutation continually generates variation in genome content and structure– Raw material for natural selection– Potential for non-adaptive evolution

• Function is too blunt a concept for genome evolution

The genome as phenotype

• Evolutionary biologists study both pattern and process, or mechanism

• What unexpected patterns do we see– Within genomes?– By comparisons among them?

• What can we infer about proximal and ultimate mechanisms by– Catching evolution red-handed?– Testing genome evolutionary models?

Three open questions

1. How rapidly do the expression profiles of duplicated genes diverge?

2. What is the extent of polymorphism in gene order within species?

3. What is responsible for physical clusters of co-expressed genes?

Three open questions

1. How rapidly do the expression profiles of duplicated genes diverge?

2. What is the extent of polymorphism in gene order within species?

3. What is responsible for physical clusters of co-expressed genes?

Arabidopsis MPSS

• With Blake Meyers and Barry Kesner• Massively parallel signature sequencing

– Bead based expression monitoring– Estimates for nearly all genes– Estimates are good even at low expression

• Sample eight tissues from Arabidopsis thaliana

• Couple with genome-wide phylogenetic analysis of duplicated genes

Measuring expression distance

Tissue 1

Tissue 2

Tissue 3

Measuring duplication age

Athal1 Athal2 Rice

T1

T2

Divergence between duplicates

Age of duplication

Exp

ress

ion

dist

ance

Does the pattern differ for

• Tandem duplicates?

• Transposed duplicates?

• Polyploidy remnants?

Three open questions

1. How rapidly do the expression profiles of duplicated genes diverge?

2. What is the extent of polymorphism in gene order within species?

3. What is responsible for physical clusters of co-expressed genes?

Gene order polymorphism

• With Jason Lieb and Jennifer Kriss• Using high-throughput methods to compare

gene order in two yeast strains• One strain known to lack 7 ORFs present in

the reference genome• The key ingredients

– Comparative genomic hybridization to a whole-genome chip

– Whole-genome genotyping in multiple haploid recombinants from a cross between the two strains

Comparative genomic hybridization on a chip

Detecting gene transposition

• Genes that are in different chromosomal positions in the two strains will appear as deletions and amplifications in the recombinant progeny

Stay tuned!

Three open questions

1. How rapidly do the expression profiles of duplicated genes diverge?

2. What is the extent of polymorphism in gene order within species?

3. What is responsible for physical clusters of co-expressed genes?

Natural selection and clusters of co-expressed genes

• With Jianhua Hu

• Neighboring genes with similar expression profiles are more common than expected in C. elegans

• Two classes of explanation– Adaptive: more efficient, less error-prone– Maladaptive: due to recent transpositions

Selection and recombination

• In regions of low recombination– deleterious mutations can hitch-hike to high

frequency along with favorable ones– favorable mutations are kept at low frequency by

linkage to deleterious ones

• The effectiveness of natural selection is directly related to recombination rate

• Are clusters found in regions of high recombination (adaptive) or low (maladaptive)?

Measuring the co-expression of neighboring genes

d1 d2 d4d3

X

• Measure the distance in expression space from central gene to each neighbor

• If neighboring genes are co-expressed, the average of these distances will be smaller than for non-neighboring genes

Sizes of co-expressed clusters

Measuring recombination rate

Expression distance and recombination rate

Spearman = -0.07p = 0.0005

Opportunities abound!

• The field of genome evolution is a vast and open playing field– Gads of genome data– Some seriously cool experimental

techniques– A golden age in bioinformatics and

molecular evolution