Dalarna University's logo and link to the university's website

The study of evolutionary dynamics requires a comprehensive understanding of how genetic interactions shape adaptation, particularly through the interplay of epistasis, pleiotropy, robustness, and evolvability. Despite significant advances in evolutionary modeling, existing frameworks often oversimplify genetic architectures, assuming fixed interactions and static environments. However, epistasis and pleiotropy are not static properties; they evolve over time, continuously reshaping the adaptive landscape. This dissertation addresses these limitations by developing a computational framework that extends the classic NK fitness landscape model through indirect encoding. This key innovation decouples epistasis and pleiotropy from landscape ruggedness, allowing these fundamental properties to evolve and to be studied independently. To systematically quantify genetic interactions, this research uses an Interaction Matrix (IM) as a tool, which captures the strength and distribution of genetic dependencies. This enables a structured approach to measuring epistasis (the degree of gene-gene interaction) and pleiotropy (the extent to which a single gene influences multiple traits). Furthermore, the study integrates Gene Regulatory Networks (GRNs) to explore how these interactions change over evolutionary time, providing insights into the role of genetic sparsity in shaping adaptive potential. The key question in evolutionary biology is how epistasis and pleiotropy influence the balance between robustness and evolvability. While robustness–the ability to maintain function despite mutations–ensures stability, evolvability–the capacity to generate beneficial variation–is crucial for long-term adaptation. The interplay between these properties is complex and widely debated. To explore this interplay, I focus on the Survival of the Flattest (SoF) phenomenon, where populations in high-mutation environments favor flatter fitness peaks over higher narrower peaks. This raises fundamental questions: How can populations remain on higher fitness peaks for extended periods? What mechanisms enable successful transitions between peaks, avoiding the SoF trap? In what scenarios does evolvability provide an advantage, and when is robustness more beneficial? I use the SoF model as a framework to investigate the role of epistasis and pleiotropy in this context. This dissertation provides new insights into these evolutionary challenges, offering a refined perspective on how populations navigate rugged landscapes, escape local fitness traps, and dynamically balance stability with adaptability. To further investigate these evolutionary dynamics, a dynamic fitness landscape is developed, allowing the study of how populations respond to changing environments. Unlike static models, the experimenter sets the velocity at which the landscape changes independently from other parameters, such as the ruggedness or the mutation rate. By varying key parameters such as ruggedness (K) and environmental change rate (V), the study examines their effects on genotype-phenotype (GP) mapping. To further validate the findings, real-world datasets were incorporated, bridging the gap between theoretical models and empirical data. The findings have broad implications beyond evolutionary biology, offering valuable insights for genetic engineering, synthetic biology, and evolutionary computation, where understanding and harnessing genetic interactions can drive innovations in biotechnology and artificial life systems. By providing a refined approach to studying how genetic architectures evolve and persist over time, this work lays a foundation for future research into the fundamental principles governing adaptive complexity. All complex systems are composed of interacting components, inherently shaped by epistasis and pleiotropy. While our model is not a literal replication of any single system, it captures these fundamental properties, allowing us to probe general principles that can illuminate the behavior of complex, real-world evolutionary dynamics.