Ernst Haeckel, 1879

From qualitative to quantitative

What are the relevant parameters?
How does adaptation and diversity depend on parameters?
How repeatable is evolution?
How predictable is evolution?
How gradual is evolution?

Population genetics & evolutionary dynamics

evolutionary processes ↔ trees ↔ genetic diversity

Concepts and typical assumptions -- 20th century

Assumptions:

The world doesn't change!
Mutations happen one at a time!
Typical events dominate!

Fitness landscapes

Diffusion theory (Kimura)

$\frac{\partial P(x,t)}{\partial t} = \frac{\partial}{\partial x}\left[-sx(1-x)+\frac{\partial}{\partial x}\frac{x(1-x)}{N}\right]P(x,t)$

Stochastic dynamics of a single locus
Difficult to generalize to many loci

Coalescence theory (Kingman)

Backwards in time model
Predictions for diversity at many loci
Difficult to generalize to non-neutral variation

Development of sequencing technologies

We can now sequence...

thousands of bacterial isolates
thousands of single cells
populations of viruses, bacteria or flies
diverse ecosystems

Experimental evolution -- Lenski experiment

Rich Lenski, Ben Good, Michael Desai et al

Influenza A/H3N2

Influenza viruses evolve to avoid human immunity
Vaccines need frequent updates

Rapidly evolving RNA viruses -- HIV

Many adaptive mutations → mutation supply not limiting
Many competing clones → complex dynamics
Environments are constantly changing

silouhette: clipartfest.com, Richman et al, 2003.

Clonal interference and traveling waves

Gerrish & Lenksi; Desai & Fisher; Kessler & Levine

Traveling wave models of adaptation

Speed of adaptation is logarithmic in population size
Environment (fitness landscape), not mutation supply, determines adaptation
Different models have universal emerging properties
Dispersal matters: well mixed populations different from dynamics in one or two dimensions

Desai & Fisher, Genetics

Bolthausen-Sznitman Coalescent

$T_c \sim \log N \quad\mathrm{instead\, of}\quad \sim N$

$\mathrm{rare\,jackpot\,events\,}P(n) \sim n^{-2}$

Brunet and Derrida, PRE, 2007; RN, Hallatschek, PNAS, 2013; Desai, Walczak, Fisher, Genetics, 2013

Testable predictions -- site frequency spectra

Kosheleva & Desai, Genetics, 2014; RN, Hallatschek, PNAS, 2013

Site frequency spectra observed in HIV populations

Zanini et al, eLife, 2015

Repeatability -- selecting 100 E. coli lines to live at 42C by Tenaillon et al

Extensive parallelism at the level of genes and pathways

Alternative adaptive path to heat adaptation

Drug resistance: repeated evolution of the same mutations is typical
Less well defined selection targets: repeatability is restricted to genes or pathways
Selection on quantitative traits: standing diversity determines response
Little success in a priori prediction of adaptive pathways

Predicting future populations

General idea

Future populations descend from present day high fitness individuals
Identify current high fitness individuals

Predicting influenza

Approach 1: leverage our understanding of trees of adapting populations
(RN, Russell, Shraiman)
Approach 2: exploit historical patterns indicative of recent adaptation
(Luksza and Lässig)
These methods now inform the WHO vaccine strain selection process.

Conclusions, limitations, questions

Good understanding of adapting populations under directional selection
Sequencing is easy, structure and function are hard
Limited ability to predict the nature of adaptations
Experimental evolution = artificial environments
→ atypical adaptation, dominated by loss of function mutations.
Is this representative? Better/other systems?
Observable pathogen evolution → Coevolution.
Is this a more general setting (or limited RNA to viruses)?
Beyond one pathogen/one host?
Ecology gets in the way → coevolution, niche construction, variable environments
How does adaptation interact with ecology and intra-species partitioning?
Can we ever separate the environment from adaptation?
How does short term adaptation extrapolate to processes on longer time scales?

Recent progress in understanding evolutionary dynamics