Learning about evolution from pathogen sequence data


Richard Neher
Biozentrum, University of Basel


slides at neherlab.org/201805_InHouse.html

Ernst Haeckel, 1879

A quantitative understanding of evolution?

  • How repeatable is evolution?
  • How gradual is evolution?
  • How predictable is evolution?
  • What are the relevant parameters?
    (population size, mutation rates etc)
  • How do the dynamics depend on parameters?

Challenges

  • Most data are static snap shots.
  • Genotype-phenotype map is largely unknown.
  • Ecology and environments are complex and variable.

Experimental evolution -- Lenski experiment

Experiment started 1927, one drop every 10 years. wikipedia.org
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


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

Evolution of HIV


  • Chimp → human transmission around 1900 gave rise to HIV-1 group M
  • ~100 million infected people since
  • subtypes differ at 10-20% of their genome
  • HIV-1 evolves ~0.1% per year
image: Sharp and Hahn, CSH Persp. Med.

HIV infection

  • $10^8$ cells are infected every day
  • the virus repeatedly escapes immune recognition
  • integrates into T-cells as latent provirus
image: wikipedia

HIV acknowledgments

  • Fabio Zanini
  • Jan Albert
  • Johanna Brodin
  • Christa Lanz
  • Göran Bratt
  • Lina Thebo
  • Vadim Puller

HIV-1 evolution within one individual



silouhette: clipartfest.com, Zanini at al, 2015. Collaboration with Jan Albert and his group

HIV-1 sequencing before and after therapy

Zanini et al, eLife, 2015; Brodin et al, eLife, 2016. Collaboration with the group of Jan Albert

Population sequencing to track all mutations above 1%

Zanini et al, eLife, 2015; antibody data from Richman et al, 2003

Diversity and rates of change

  • envelope changes fastest, enzymes slowest
  • identical rate of synonymous evolution
  • diversity saturates where evolution is fast
  • synonymous mutations stay at low frequency
Zanini et al, eLife, 2015

Estimating the date of infection from diversity data

  • diversity at 3rd positions increases almost linearly in time
  • can be used to predict date of infection
  • critical to estimate incidence
  • Multiple founder viruses cause over estimation
  • Degraded samples cause under estimation
Puller et al, PLoS Comp Bio, 2017

Inference of fitness costs

  • mutation away from preferred state with rate $\mu$
  • selection against non-preferred state with strength $s$
  • variant frequency dynamics: $\frac{d x}{dt} = \mu -s x $
  • equilibrium frequency: $\bar{x} = \mu/s $
  • fitness cost: $s = \mu/\bar{x}$

Fitness landscape of HIV-1

Zanini et al, Virus Evolution, 2017

Selection on RNA structures and regulatory sites

  • red: mutations that don't change protein sequence
  • blue: all mutations
Zanini et al, Virus Evolution, 2017

Influenza A/H3N2 virus evolution

Joint work with....

  • Boris Shraiman
  • Colin Russell
  • Trevor Bedford

nextflu.org

joint work with Trevor Bedford & his lab

Prediction of the dominating H3N2 influenza strain

RN, Russell, Shraiman, eLife, 2014

nextstrain.org

  • Trevor Bedford
  • Colin Megill
  • Pavel Sagulenko
  • Sidney Bell
  • James Hadfield
  • Wei Ding
  • Emma Hodcroft

nextstrain.org

joint work with Trevor Bedford & his lab

nextstrain.org

  • integrate data from many different sources
  • analyze those data in near real time
  • disseminate results in an intuitive yet informative way
  • provide actionable insights

What about more complicated things than viruses?

nextTB: real-time molecular epidemiology of TB

Collaboration with Sebastian Gagneux and colleages at the STPH

  • 10s of thousand MTB genomes have been sequenced
  • Elucidate transmission routes at the local and global level
  • Integrate with drug resistance data

Pan-genomes of bacteria

  • much larger genomes
  • vertical and horizontal transmission
  • gene gain and loss
  • genome rearrangements
  • variation of divergence along the genome
panX by Wei Ding

Pan-genome statistics and filters

Species trees and gene trees

Genome assembly with short reads

  • 10s of millions of short reads (<500bp)
  • Too short to bridge repetitive elements
  • → assemblies are fragmented into 100s of "contigs"

(really terrible example)

Images: illumina.com, github.com/rrwick

Long-read sequencing

Acknowledgements

Group

  • Fabio Zanini
  • Pavel Sagulenko
  • Vadim Puller
  • Wei Ding
  • Sanda Dejanic
  • Emma Hodcroft
  • Nicholas Noll
  • Eric Ulrich

Karolinska Institute

  • Jan Albert
  • Lina Thebo
  • Johanna Brodin

FHCRC

  • Trevor Bedford
  • James Hadfield
  • Sidney Bell
  • Colin Megill

Swiss TPH

  • Sebastian Gagneux
  • Chloé Loiseau
  • Fabrizio Menardo

USB

  • Adrian Egli
  • Daniela Lang

BZ

  • Leo Faletti
  • sciCore
  • workshops
  • Boris Shraiman (UCSB)
  • Colin Russell (AMC)
  • Oskar Hallatschek (UCB)