Computational Methods for Mapping Genetic Interactions in Human Cells

点击翻译按钮获取中文摘要

Project Summary Despite our increasing capabilities to efficiently measure human genomes, we still face major challenges in interpreting them. Even for diseases or other traits that have a strong genetic component, we are generally still unable to accurately predict disease status from genome sequence. One major reason for this gap is that we still do not understand the rules for how variants at different loci in the genome combine to affect an organism. Genes do not act independently, and the effect of one genetic change can depend greatly on the presence of other variants in a genome. Over the past two decades, extensive work has been carried out in model organisms such as yeast to explore the basic principles that govern how genes interact to cause phenotypes. In particular, several efforts have systematically introduced combinations of precisely engineered mutations on a global scale and measured their effects on cells. This work has demonstrated that systematic combinatorial genome perturbation can be a powerful strategy to understand how a genome is functionally organized and can precisely elucidate the functional role of the specific components. While technical challenges have previously limited similar endeavors in higher organisms, new disruptive CRISPR/Cas9-based genome editing technology now makes this powerful combinatorial mutation approach possible on the human genome. However, although the experimental technology now exists to systematically manipulate the human genome on a genome-wide scale, we still lack the computational approaches necessary for interpreting the resulting data. The specific objective of this proposal is to develop new computational methods that directly support the systematic mapping and analysis of genetic interactions in human cells based on CRISPR/Cas9 technology. We will accomplish our objective by focusing on three specific aims: (1) develop computational models for measuring quantitative genetic interactions from genome-wide CRISPR/Cas9 screens in human cells, (2) develop algorithms for optimal query selection that enable efficient strategies for mapping human genetic interactions, and (3) apply optimal screen selection algorithm to develop a scalable screening platform for functional profiling of chemical perturbations and genetic variants. The proposed research is innovative in that it bridges concepts established over a decade of work on genetic interactions in yeast with the latest developments in genome editing technology. Our proposed work will establish new, robust computational tools that will be broadly applicable to large-scale CRISPR/Cas9- based screening efforts.