Phenotype inference in an Escherichia coli strain panel

摘要: Understanding how genetic variation contributes to phenotypic differences is a fundamental question in biology. Combining high-throughput gene function assays with mechanistic models of the impact variants promising alternative genome-wide association studies. Here we have assembled large panel 696 Escherichia coli strains, which genotyped and measured their profile across 214 growth conditions. We integrated variant effect predictors derive gene-level probabilities loss for every all strains. Finally, combined these information on conditional essentiality reference K-12 strain compute defects each strain. Not only could reliably predict up 38% tested conditions, but also directly identify causal that were validated through complementation assays. Our work demonstrates power forward predictive possibility precision interventions.