Copy number variants (CNVs) are a pervasive source of genetic variation and evolutionary potential, but the dynamics and diversity of CNVs within evolving populations remain unclear. Long-term evolution experiments in chemostats provide an ideal system for studying the molecular processes underlying CNV formation and the temporal dynamics with which they are generated, selected, and maintained. Here, we developed a fluorescent CNV reporter to detect de novo gene amplifications and deletions in individual cells. We used the CNV reporter in Saccharomyces cerevisiae to study CNV formation at the GAP1 locus, which encodes the general amino acid permease, in different nutrient-limited chemostat conditions. We find that under strong selection, GAP1 CNVs are repeatedly generated and selected during the early stages of adaptive evolution, resulting in predictable dynamics. Molecular characterization of CNV-containing lineages shows that the CNV reporter detects different classes of CNVs, including aneuploidies, nonreciprocal translocations, tandem duplications, and complex CNVs. Despite GAP1’s proximity to repeat sequences that facilitate intrachromosomal recombination, breakpoint analysis revealed that short inverted repeat sequences mediate formation of at least 50% of GAP1 CNVs. Inverted repeat sequences are also found at breakpoints at the DUR3 locus, where CNVs are selected in urea-limited chemostats. Analysis of 28 CNV breakpoints indicates that inverted repeats are typically 8 nucleotides in length and separated by 40 bases. The features of these CNVs are consistent with origin-dependent inverted-repeat amplification (ODIRA), suggesting that replication-based mechanisms of CNV formation may be a common source of gene amplification. We combined the CNV reporter with barcode lineage tracking and found that 102–104 independent CNV-containing lineages initially compete within populations, resulting in extreme clonal interference. However, only a small number (18–21) of CNV lineages ever constitute more than 1% of the CNV subpopulation, and as selection progresses, the diversity of CNV lineages declines. Our study introduces a novel means of studying CNVs in heterogeneous cell populations and provides insight into their dynamics, diversity, and formation mechanisms in the context of adaptive evolution.

The rate of adaptive evolution depends on the rate at which beneficial mutations are introduced into a population and the fitness effects of those mutations. The rate of beneficial mutations and their expected fitness effects is often difficult to empirically quantify. As these 2 parameters determine the pace of evolutionary change in a population, the dynamics of adaptive evolution may enable inference of their values. Copy number variants (CNVs) are a pervasive source of heritable variation that can facilitate rapid adaptive evolution. Previously, we developed a locus-specific fluorescent CNV reporter to quantify CNV dynamics in evolving populations maintained in nutrient-limiting conditions using chemostats. Here, we use CNV adaptation dynamics to estimate the rate at which beneficial CNVs are introduced through de novo mutation and their fitness effects using simulation-based likelihood–free inference approaches. We tested the suitability of 2 evolutionary models: a standard Wright–Fisher model and a chemostat model. We evaluated 2 likelihood-free inference algorithms: the well-established Approximate Bayesian Computation with Sequential Monte Carlo (ABC-SMC) algorithm, and the recently developed Neural Posterior Estimation (NPE) algorithm, which applies an artificial neural network to directly estimate the posterior distribution. By systematically evaluating the suitability of different inference methods and models, we show that NPE has several advantages over ABC-SMC and that a Wright–Fisher evolutionary model suffices in most cases. Using our validated inference framework, we estimate the CNV formation rate at the GAP1 locus in the yeast Saccharomyces cerevisiae to be 10−4.7 to 10−4 CNVs per cell division and a fitness coefficient of 0.04 to 0.1 per generation for GAP1 CNVs in glutamine-limited chemostats. We experimentally validated our inference-based estimates using 2 distinct experimental methods—barcode lineage tracking and pairwise fitness assays—which provide independent confirmation of the accuracy of our approach. Our results are consistent with a beneficial CNV supply rate that is 10-fold greater than the estimated rates of beneficial single-nucleotide mutations, explaining the outsized importance of CNVs in rapid adaptive evolution. More generally, our study demonstrates the utility of novel neural network–based likelihood–free inference methods for inferring the rates and effects of evolutionary processes from empirical data with possible applications ranging from tumor to viral evolution.

Copy number variants (CNVs) are a pervasive, but understudied source of genetic variation and evolutionary potential. Long-term evolution experiments in chemostats provide an ideal system for studying the molecular processes underlying CNV formation and the temporal dynamics of de novo CNVs. Here, we developed a fluorescent reporter to monitor gene amplifications and deletions at a specific locus with single-cell resolution. Using a CNV reporter in nitrogen-limited chemostats, we find that GAP1 CNVs are repeatedly generated and selected during the early stages of adaptive evolution resulting in predictable dynamics of CNV selection. However, subsequent diversification of populations defines a second phase of evolutionary dynamics that cannot be predicted. Using whole genome sequencing, we identified a variety of GAP1 CNVs that vary in size and copy number. Despite GAP1 's proximity to tandem repeats that facilitate intrachromosomal recombination, we find that non-allelic homologous recombination (NAHR) between flanking tandem repeats occurs infrequently. Rather, breakpoint characterization revealed that for at least 50% of GAP1 CNVs, origin-dependent inverted-repeat amplification (ODIRA), a DNA replication mediated process, is the likely mechanism. We also find evidence that ODIRA generates DUR3 CNVs, indicating that it may be a common mechanism of gene amplification. We combined the CNV reporter with barcode lineage tracking and found that 10 3 -10 4 independent CNV-containing lineages initially compete within populations, which results in extreme clonal interference. Our study introduces a novel means of studying CNVs in heterogeneous cell populations and provides insight into the underlying dynamics of CNVs in evolution.condition [32,33] . A high rate of CNV formation suggests that multiple, independent CNV-containing lineages may compete during adaptive evolution resulting in clonal interference, which is characteristic of large, evolving populations [29,[34][35][36] . However, the extent of clonal interference among CNV-containing lineages is unknown.The general amino acid permease, GAP1 , is ideally suited to studying the role of CNVs in adaptive evolution. GAP1 encodes a high-affinity transporter for all naturally occurring amino acids and analogues, and it is highly expressed in nitrogen-poor conditions [37,38] . We have previously shown that two classes of CNVs are selected at the GAP1 locus in S. cerevisiae : amplification alleles in glutamine and glutamate-limited chemostats and deletion alleles in ureaand allantoin-limited chemostats [24,25] . GAP1 CNVs are also found in natural populations.Multiple, tandem copies of GAP1 have been identified in wild populations of the nectar yeast, Metschnikowia reukaufii , which result in a competitive advantage over other microbes when amino acids are scarce [39] . As a frequent target of selection in adverse environments in both experimental and natural populations, GAP1 is a model locus for studying the dynamics and mechanisms underlying both gene amplification a...

Large-scale genomic changes, including copy number variations (CNVs), are frequently observed in long-term evolution experiments (LTEEs). We have previously reported the detection of recurrent CNVs in Saccharomyces cerevisiae populations adapting to glutamine-limited conditions over hundreds of generations. Here, we present the whole-genome sequencing (WGS) assemblies of 7 LTEE strains and their ancestor.

The rate of adaptive evolution depends on the rate at which beneficial mutations are introduced into a population and the fitness effects of those mutations. The rate of beneficial mutations and their expected fitness effects is often difficult to empirically quantify. As these two parameters determine the pace of evolutionary change in a population, the dynamics of adaptive evolution may enable inference of their values. Copy number variants (CNVs) are a pervasive source of heritable variation that can facilitate rapid adaptive evolution. Previously, we developed a locus-specific fluorescent CNV reporter to quantify CNV dynamics in evolving populations maintained in nutrient-limiting conditions using chemostats. Here, we use the observed CNV adaptation dynamics to estimate the rate at which beneficial CNVs are introduced through de novo mutation and their fitness effects using simulation-based Bayesian likelihood-free inference approaches. We tested the suitability of two evolutionary models: a standard Wright-Fisher model and a chemostat growth model. We evaluated two likelihood-free inference algorithms: the well-established Approximate Bayesian Computation with Sequential Monte Carlo (ABC-SMC) algorithm, and the recently developed Neural Posterior Estimation (NPE) algorithm, which applies an artificial neural network to directly estimate the posterior distribution. By systematically evaluating the suitability of different inference methods and models we show that NPE has several advantages over ABC-SMC and that a Wright-Fisher evolutionary model suffices in most cases. Using our validated inference framework, we estimate the CNV formation rate at the GAP1 locus in yeast as 10-4.7 -10-4 per cell division, and a selection coefficient of 0.04 - 0.1 per generation for GAP1 CNVs in glutamine-limited chemostats. We experimentally validated our estimates using barcode lineage tracking and pairwise fitness assays. Our results are consistent with a high beneficial CNV supply rate that is 10-fold greater than the estimated rates of beneficial single-nucleotide mutations, explaining their outsized importance in rapid adaptive evolution. More generally, our study demonstrates the utility of novel simulation-based likelihood-free inference methods for inferring the rates and effects of evolutionary processes from empirical data.