Single locus theory indicates that selfing species are more able than outcrossing ones to fix emerging recessive beneficial mutations, as they are not masked as heterozygotes. However, partially selfing organisms suffer from relaxed recombination, which reduces overall selection efficiency. Although the effect of linked deleterious alleles on adaptation has previously been studied, the extent to which multiple adaptations interfere in partially selfing organisms is currently unknown. We derive branching-process models to quantify the extent that emergence of a second beneficial allele is obstructed by an existing selective sweep. We consider both the potential loss of the second beneficial mutation if it has a weaker advantage than the first sweep (the `stochastic interference effect’), and also the potential replacement of the first sweep if the second mutant is fitter (`replacement effect’). Overall, the stochastic interference effect has a larger impact on preventing fixation of both adaptive alleles in highly selfing organisms, but the replacement effect can be stronger with multiple mutations. Interference has two opposing effects on Haldane’s Sieve. First, recessive mutants are disproportionally likely to be lost, so it is more likely that only dominant mutations will emerge in outcrossers. Second, with frequent rates of adaptive evolution, outcrossing organisms are more able to fix weak beneficial mutations of any dominance value, contrary to the predictions of Haldane’s Sieve. Our analysis shows that even under low rates of adaptive mutation, interference can be sufficiently strong to greatly limit adaptation in selfing organisms.

Viruses are the simplest replicating units, characterized by a limited number of coding genes and an exceptionally high rate of overlapping genes. We sought a unified explanation for the evolutionary constraints that govern genome sizes, gene overlapping and capsid properties. We performed an unbiased statistical analysis over the ~100 known viral families, and came to refute widespread assumptions regarding viral evolution. We found that the volume utilization of viral capsids is often low, and greatly varies among families. Most notably, we show that the total amount of gene overlapping is tightly bounded. Although viruses expand three orders of magnitude in genome length, their absolute amount of gene overlapping almost never exceeds 1500 nucleotides, and mostly confined to <4 significant overlapping instances. Our results argue against the common theory by which gene overlapping is driven by a necessity of viruses to compress their genome. Instead, we support the notion that overlapping has a role in gene novelty and evolution exploration.

The origin of the germline-soma distinction in metazoans is a fundamental unsolved question. Somatic gametogenesis in sessile sponges and corals contrasts starkly with early germline sequestration in bilaterians with higher energy requirements, elaborate body plans and fast evolution of small mitochondrial genomes. We develop an evolutionary model to investigate whether selection for mitochondrial quality can drive germline evolution. In basal metazoans with low mutation rates, somatic gametogenesis optimizes gamete quality through segregation of mitochondrial mutations in multiple cell divisions. The need to maintain mitochondrial quality in somatic tissues explains the evolution of oogamy and male-female gamete specialization. Higher mitochondrial mutation rates promote early sequestration of a dedicated germline, permitting complex developmental processes. Rising oxygen drove germline evolution in motile bilaterians, igniting the Cambrian explosion.

1) Micro-evolutionary predictions are complicated by ecological feedbacks like density dependence, while ecological predictions can be complicated by evolutionary change. A widely used approach in micro-evolution, quantitative genetics, struggles to incorporate ecological processes into predictive models, while structured population modelling, a tool widely used in ecology, rarely incorporates evolution explicitly. 2) In this paper we develop a flexible, general framework that links quantitative genetics and structured population models. We use the quantitative genetic approach to write down the phenotype as an additive map. We then construct integral projection models for each component of the phenotype. The dynamics of the distribution of the phenotype are generated by combining distributions of each of its components. Population projection models can be formulated on per generation or on shorter time steps. 3) We introduce the framework before developing example models with parameters chosen to exhibit specific dynamics. These models reveal (i) how evolution of a phenotype can cause populations to move from one dynamical regime to another (e.g. from stationarity to cycles), (ii) how additive genetic variances and covariances (the G matrix) are expected to evolve over multiple generations, (iii) how changing heritability with age can maintain additive genetic variation in the face of selection and (iii) life history, population dynamics, phenotypic characters and parameters in ecological models will change as adaptation occurs. 4) Our approach unifies population ecology and evolutionary biology providing a framework allowing a very wide range of questions to be addressed. The next step is to apply the approach to a variety of laboratory and field systems. Once this is done we will have a much deeper understanding of eco-evolutionary dynamics and feedbacks.

Genomic instability causes cancers to acquire hundreds to thousands of mutations and chromosomal alterations during their somatic evolution. Most of these mutations and alterations are termed passengers because they do not confer cancer phenotypes. Evolutionary simulations and cancer genomic studies suggested that mildly-deleterious passengers accumulate, collectively slow cancer progression, reduce the fitness of cancer cells and enhance the effects of therapeutics. However, these effects of passengers and their impact on clinical variables remain limited to genomic analysis. Here, to assess passengers’ effect on cell fitness and cancer, we specifically introduced increasing passenger loads into human cell lines and mouse models. We found that passenger load dramatically reduced cancer cell’s fitness in every model investigated. Passengers’ average fitness cost of ~3% per MB, indicates that genomic instability in cancer in patients can slow tumor growth and prevent metastatic progression. We conclude that genomic instability in cancer is a double-edged sword: it accelerates the accumulation of adaptive drivers, yet incurs a harmful passenger load that can outweigh drivers’ benefit. Passenger load could be a useful biomarker for tumor aggressiveness and response to mutagenic or passenger-exacerbating therapies, including anti-tumor immunity.

Many long noncoding RNAs (lncRNAs) can regulate chromatin states, but the evolutionary origin and dynamics driving lncRNA-genome interactions are unclear. We developed an integrative strategy that identifies lncRNA orthologs in different species despite limited sequence similarity that is applicable to fly and mammalian lncRNAs. Analysis of the roX lncRNAs, which are essential for dosage compensation of the single X-chromosome in Drosophila males, revealed 47 new roX orthologs in diverse Drosophilid species across ~40 million years of evolution. Genetic rescue by roX orthologs and engineered synthetic lncRNAs showed that evolutionary maintenance of focal structural repeats mediates roX function. Genomic occupancy maps of roX RNAs in four species revealed rapid turnover of individual binding sites but conservation within nearby chromosomal neighborhoods. Many new roX binding sites evolved from DNA encoding a pre-existing RNA splicing signal, effectively linking dosage compensation to transcribed genes. Thus, evolutionary analysis illuminates the principles for the birth and death of lncRNAs and their genomic targets.