About

Sergey Kryazhimskiy is an Associate Professor with a PhD in Applied and Computational Mathematics from Princeton University. His research centers on understanding, predicting, and controlling the evolution and adaptation of microbes to their environments. He approaches evolution by natural selection as a fundamental property of life and aims to uncover the mechanisms by which microbial populations evolve over time. His work integrates mathematical and computational methods to study evolutionary systems biology, focusing on microbial adaptation and evolutionary dynamics.

Research topics

• Genetics
• Evolutionary biology
• Biology
• Ecology
• Neuroscience
• Engineering
• Cell biology
• Biochemical engineering
• Computational biology

Selected publications

• Upper bound on the mutational burden imposed by a CRISPR-Cas9 gene-drive element

  G3 Genes Genomes Genetics · 2026-02-18

  articleOpen accessSenior author

  Homing-based CRISPR-Cas9 gene drives (CCGDs) are powerful tools for genetic control of wild populations, with applications from disease eradication to species conservation. However, Cas9 alone and in a complex with guide RNA can cause double-stranded DNA breaks at off-target sites, which could increase the mutational load and lead to unintended loss-of-heterozygosity (LOH) events. These undesired effects raise potential concerns about the long-term evolutionary safety of CCGDs, but the magnitude of these effects is unknown. To measure how the presence of a CCGD or a Cas9 alone in the genome affects the rates of LOH events and de novo mutations, we carried out a mutation accumulation experiment in yeast Saccharomyces cerevisiae. We found no detectable effects on the genome-wide rates of mutations or LOH events. Our power calculations suggest that CCGD or Cas9 affect these rates by less than 30%, which is much less than natural variation for these traits in yeast. A more detailed examination shows that CCGD or Cas9 may alter the lengths and genomic distributions of LOH events, but the statistical support for these effects is weak. Thus, our results demonstrate that CCGDs impose at most a weak additional mutational burden in the yeast model. Although mutagenic effects of gene drives need to be further evaluated in other systems, our results add credence to the proposition that the evolutionary risks posed by well-designed gene drives may be acceptable.

  PublisherDOI

• Module-Selection Balance in the Evolution of Modular Organisms

  bioRxiv (Cold Spring Harbor Laboratory) · 2026-04-03

  articleOpen accessSenior authorCorresponding

  Abstract The architecture of the genotype-phenotype-fitness map (GPFM) is a key determinant of evolutionary dynamics. One salient feature of biological GPFMs is variational modularity, where each mutation affects only a small subset of functional traits. Variational modularity may constrain the dynamics of trait evolution, but these constraints are not well understood. Here, we use several extensions of the Fisher’s geometric model with two functional traits to investigate these constrains. We find that on GPFMs with universal pleiotropy, populations evolve along the fitness gradient, which implies that the trait under stronger selection is optimized exponentially faster than the trait under weaker selection. In contrast, on modular GPFMs, populations approach a quasi-steady state that we term a “module-selection balance” where both traits improve at the same rate and their ratio remains constant. We demonstrate that the existence of a module-selection balance is robust with respect to the details of evolutionary dynamics and GPFMs themselves, as long as they are variationally modular. Our theory predicts that variationally modular organisms should exhibit stereotypical bi-phasic dynamics of genome evolution, especially in the strong clonal interference regime, and we find support for this prediction in metagenomic data from Lenski’s long-term evolution experiment in bacterium Escherichia coli . We propose that module-selection balance is an inherent feature of variationally modular GPFMs, which imposes an important constraint on long-term trait evolution.

  PublisherOA PDFDOI

• Putting BASIL in a BLT: A Bayesian filtering method for estimating the fitness effects of nascent adaptive mutations

  PLoS Computational Biology · 2026-02-27

  articleOpen accessSenior author

  The distribution of fitness effects (DFE) of new beneficial mutations is a key quantity that dictates the dynamics of adaptation. The barcode lineage tracking (BLT) approach is an important advance toward measuring DFEs. BLT experiments enable researchers to track the frequencies of ~105 barcoded lineages in large microbial populations and detect up to thousands of nascent beneficial mutations in a single experiment. However, reliably identifying adapted lineages and estimating the fitness effects of driver mutations remains a challenge because lineage dynamics are subject to demographic and measurement noise and competition with other lineages. We show that the commonly used Levy-Blundell method for analyzing BLT data and its improved version FitMut2 can produce biased fitness estimates, particularly if selection is strong. To address this problem, we develop a new method called BASIL (BAyesian Selection Inference for Lineage tracking data), which dynamically updates the belief distribution of each lineage's fitness and size based on the number of barcode reads. We calibrate BASIL's model of noise with new experimental data and find that noise variance scales non-linearly with lineage abundance. We test how BASIL and FitMut2 perform on simulated data and on down-sampled data from the original BLT data by Levy et al and find that BASIL is both more robust and more accurate than FitMut2. Our work paves the way for a systematic inference of the distribution of fitness effects of new beneficial mutations from BLT experiments in a variety of scenarios.

  PublisherDOI

• Improved genotype inference reveals cis- and trans-driven variation in the loss-of-heterozygosity rates in yeast

  Genetics · 2025-12-22 · 1 citations

  articleOpen accessSenior author

  Loss-of-heterozygosity (LOH) events are an important source of genetic variation in diploids and are implicated in cancer. LOH-event rates vary across the genome and across genetic backgrounds, but our understanding of this variation is incomplete. State-of-the-art measurements of LOH rates are obtained from mutation accumulation (MA) experiments in heterozygous hybrids, mainly in yeast Saccharomyces cerevisiae. These measurements hinge on the accuracy of inference of diploid genotypes from short sequencing reads. We analyzed a new large yeast MA dataset and found that the currently standard "single-reference" genotyping approach can lead to errors in LOH-rate estimates and produce spurious homolog biases. To address this problem, we developed a novel genotyping approach for MA experiments that is symmetric with respect to both homologs, removes dubious heterozygous markers, and corrects for undetected LOH events. We report revised estimates of LOH rates across 12 yeast hybrids, which differ by factors between 0.19 and 5.3 from previously published ones. Our revised estimates do not support the previously reported positive correlation between the rate of terminal LOH events and the hybrid heterozygosity. Finally, our analysis reveals that the 60-fold variation in the rates of interstitial LOH events across yeast hybrids is driven overwhelmingly by genetic factors with genome-wide (trans) effects. In contrast, the 6-fold variation in terminal LOH events is driven by both trans and local (cis) factors. Our results provide a foundation for reliable detection of LOH events and further investigations into the genetic underpinnings of LOH-rate variation.

  PublisherDOI

• Improved Genotype Inference Reveals Cis- and Trans-Driven Variation in the Loss-of-Heterozygosity Rates in Yeast

  bioRxiv (Cold Spring Harbor Laboratory) · 2025-06-30 · 1 citations

  preprintOpen accessSenior author

  Abstract Loss-of-heterozygosity (LOH) events are an important source of genetic variation in diploids and are implicated in cancer. LOH-event rates vary across the genome and across genetic backgrounds, but our understanding of this variation is in its infancy. State-of-the-art measurements of LOH rates are obtained from mutation accumulation (MA) experiments in heterozygous hybrids and hinge on the accurate inference of diploid genotypes from short sequencing reads. We analyzed a new large MA dataset in yeast Saccharomyces cerevisiae and found that the currently standard “single-reference” genotyping approach can lead to errors in LOH-rate estimates and produce spurious homolog biases. To address this problem, we develop a novel genotyping approach that is symmetric with respect to both homologs, removes dubious heterozygous markers and corrects for undetected LOH events. We report revised estimates of LOH rates across 12 yeast hybrids, which differ by factors between 0.19 and 5.3 from previously published ones. Our revised estimates do not support the previously reported positive correlation between the rate of terminal LOH events and the hybrid heterozygosity. Finally, our analysis reveals that the 60-fold variation in the rates of interstitial LOH events across yeast hybrids is driven overwhelmingly by genome-wide (trans) genetic factors. In contrast, the 6-fold variation in terminal LOH events is driven by both trans and local (cis) factors. Our results provide a foundation for reliable detection of LOH events and further investigations into the genetic underpinnings of LOH-rate variation.

  PublisherDOI

• Putting BASIL in a BLT: A Bayesian filtering method for estimating the fitness effects of nascent adaptive mutations

  bioRxiv (Cold Spring Harbor Laboratory) · 2025-04-03

  preprintOpen accessSenior authorCorresponding

  The distribution of fitness effects (DFE) of new beneficial mutations is a key quantity that dictates the dynamics of adaptation. The barcode lineage tracking (BLT) approach is an important advance toward measuring DFEs. BLT experiments enable researchers to track the frequencies of ~10 5 of barcoded lineages in large microbial populations and detect up to thousands of nascent beneficial mutations in a single experiment. However, reliably identifying adapted lineages and estimating the fitness effects of driver mutations remains a challenge because lineage dynamics are subject to demographic and measurement noise and competition with other lineages. We show that the commonly used Levy-Blundell method for analyzing BLT data and its improved version FitMut2 can produce biased fitness estimates, particularly if selection is strong. To address this problem, we develop a new method called BASIL (BAyesian Selection Inference for Lineage tracking data), which dynamically updates the belief distribution of each lineage's fitness and size based on the number of barcode reads. We calibrate BASIL's model of noise with new experimental data and find that noise variance scales non-linearly with lineage abundance. We test how BASIL and FitMut2 perform on simulated data and on down-sampled data from the original BLT data by Levy et al and find that BASIL is both more robust and more accurate than FitMut2. Our work paves the way for a systematic inference of the distribution of fitness effects of new beneficial mutations from BLT experiments in a variety of scenarios.

  PublisherOA PDFDOI

• Environment-independent distribution of mutational effects emerges from microscopic epistasis

  Science · 2024-10-03 · 29 citations

  articleOpen accessSenior authorCorresponding

  Predicting how new mutations alter phenotypes is difficult because mutational effects vary across genotypes and environments. Recently discovered global epistasis, in which the fitness effects of mutations scale with the fitness of the background genotype, can improve predictions, but how the environment modulates this scaling is unknown. We measured the fitness effects of ~100 insertion mutations in 42 strains of Saccharomyces cerevisiae in six laboratory environments and found that the global epistasis scaling is nearly invariant across environments. Instead, the environment tunes one global parameter, the background fitness at which most mutations switch sign. As a consequence, the distribution of mutational effects is predictable across genotypes and environments. Our results suggest that the effective dimensionality of genotype-to-phenotype maps across environments is surprisingly low.

  PublisherOA PDFDOI

• A simple rule for predicting function of microbial communities

  Cell · 2024-06-01 · 1 citations

  articleOpen access1st authorCorresponding
  PublisherOA PDFDOI

• :<i>The Evolution of Biological Information: How Evolution Creates Complexity, from Viruses to Brains</i>

  The Quarterly Review of Biology · 2024-11-15

  review1st authorCorresponding
  PublisherDOI

• Best Practices in Designing, Sequencing, and Identifying Random DNA Barcodes

  Journal of Molecular Evolution · 2023-01-18 · 47 citations

  reviewOpen accessSenior author

  Random DNA barcodes are a versatile tool for tracking cell lineages, with applications ranging from development to cancer to evolution. Here, we review and critically evaluate barcode designs as well as methods of barcode sequencing and initial processing of barcode data. We first demonstrate how various barcode design decisions affect data quality and propose a new design that balances all considerations that we are currently aware of. We then discuss various options for the preparation of barcode sequencing libraries, including inline indices and Unique Molecular Identifiers (UMIs). Finally, we test the performance of several established and new bioinformatic pipelines for the extraction of barcodes from raw sequencing reads and for error correction. We find that both alignment and regular expression-based approaches work well for barcode extraction, and that error-correction pipelines designed specifically for barcode data are superior to generic ones. Overall, this review will help researchers to approach their barcoding experiments in a deliberate and systematic way.

  PublisherOA PDFDOI

Recent grants

• Characterizing variation in the local structure of fitness landscapes to assess the predictability of evolution

  NIH · $1.2M · 2020–2024

Frequent coauthors

• Michael M. Desai

  Quantitative BioSciences

  29 shared

• Daniel P. Rice

  Massachusetts Institute of Technology

  24 shared

• Gabriel G. Perron

  New York University

  18 shared

• Angus Buckling

  University of Exeter

  18 shared

• Georgii A. Bazykin

  Institute for Information Transmission Problems

  18 shared

• Joshua B. Plotkin

  University of Pennsylvania

  17 shared

• Elizabeth R. Jerison

  13 shared

• Milo S. Johnson

  University of California, Berkeley

  12 shared

Labs

• SKLab @ UCSDPI

  Evolutionary Systems Biology

Education

• PhD, Applied and Computational Mathematics

  Princeton University

  2008

Awards & honors

• Burroughs Wellcome Fund Career Award at Scientific Interface