A path to personalized phenotypic prediction: unlocking the context-dependency of allelic effects

点击翻译按钮获取中文摘要

The long-term goal of my research program is to understand the biological basis for individual variation. The genetic architecture of complex traits is not a static blueprint of the phenotype as it was previously thought; rather, it is highly dynamic and context-dependent. I seek to understand how genes interact with each other and their environment to shape variation between individuals and what factors control the degree of individual variability. Technological advances have recently fueled the ascent of personal genomics and the promise of precision medicine. The success of medical genetics will depend on its capacity to personalize, however, individualized prediction is a grand challenge. When the average effect of an allele does not capture a specific allelic contribution under certain conditions (whether due to genetic background or the environment), the link between genotype and phenotype will be missed. Given such context dependency, understanding how genotypic variation influences variation in an individual's phenotype demands a shift in focus from population averages to individual effects. Globally, we are witnessing the rise of complex diseases related to dramatic changes in our daily environments. These disorders have a clear environmental basis, but they also show strong familial correlations: susceptibility to these diseases is highly heritable. Despite considerable effort and resources, we have made little progress in understanding the genetic basis of these common conditions. This highlight the need for a different approach to identify the causal genetic factors underlying disorders characterized by non-additive interactions. To date, a key limitation to address this problem has been that small sample sizes and skewed allele frequency spectrum limit the power of detecting genetic associations. We have solved this problem by creating a new community resource made of large, synthetic outbred populations. This enables us to break away from traditional, artificial and underpowered approaches that have relied on inbred strains. In parallel, we have developed a molecular and analytical pipeline allowing us to sequence thousands of single flies at high throughput with very low cost and reliable accuracy. With this new and versatile resource, we can rear thousands of genetically unique flies drawn from a common genetic pool, expose them to a range of different environments, and contrast the ensuing genetic architectures. Our inability to make progress in human genetics for diseases with strong environmental components suggests a fundamental knowledge gap that my research addresses in a powerful model system. Given that in humans there is extreme variation and stochasticity in environmental exposure, we need a predictive framework that can accommodate these individual-specific impacts. My research program paves a path to personalized phenotypic prediction by unlocking the context dependence of allelic effects.