Tracking and interpreting novel SARS-CoV-2 variants

Richard Neher
Biozentrum & SIB, University of Basel

slides at

Weekly sequence submissions

Data we analyze are contributed by scientists from all over the world

Data are shared via and curated by GISAID

Nextstrain builds for different countries

  • Started as part of the surveillance effort in Switzerland
  • Most European countries -- updated twice a week
  • Heavy focus on sequences from the last 4 weeks (but still subsampled in many cases)
  • Administrative divisions for some countries
  • Variant specific builds via CoVariants
  • We sometimes include experimental features
  • Moderate effort to add more -- get in touch if you have specific ideas or requests

Similarities to influenza virus evolution

Klingen et al, Trends in MicroBio, 2018

Continuous back and forth between experimental and computational groups.

Challenges assessing novel variants

  • Frequency dynamics are skewed by sampling biases and varying incidence
  • Rare sequences might represent large outbreaks in poorly sampled regions
  • Sudden changes in sampling strategies might look like rapidly increasing frequencies
  • Focus on specific mutations in known VoCs might miss important others
  • Laboratory characterization takes time
  • Prioritization of variants for laboratory characterization.
  • Harmonized categorization of sequences as
    • General surveillance
    • Outbreak investigations
    • Vaccine breakthroughs, reinfections, etc
    • Travel associated cases
  • Support for denser sampling across the globe
    (some places in Europe are still undersampled)

Deep mutational scanning for RBD mutations that affect AB binding and ACE2 binding

Scoring all sequences for further characterization

  • Escape in distinct RBD epitope classes
  • RBD escape in convalescent/vaccine sera
  • NTD mutations
  • Other VoC associated mutations (nsp6 del)
  • ACE2 binding
  • Furin cleavage site mutations
  • Recent growth

Incorporation of scores into phylogenetic analysis


Trevor Bedford and his lab -- terrific collaboration since 2014

especially James Hadfield, Emma Hodcroft, Ivan Aksamentov, Moira Zuber, and Tom Sibley

Allie Greaney and Jesse Bloom for discussion of the dms data

Data we analyze are contributed by scientists from all over the world

Data are shared and curated by GISAID