PROJECT SUMMARY Elucidating the genetic and environmental drivers of disease is critical to precision medicine. Advancements in in vitro human stem cell culture are accelerating disease and therapeutic research across individuals of di- verse genetic backgrounds. Cell villages are an e!cient platform, that pools populations of 10-100 genetically distinct, human stem cell lines, to study how genetic variation and environment influence a spectrum of molec- ular phenotypes. This experimental design allows researchers to apply selection and perturbation treatments to large, ancestrally diverse cell populations, while simultaneously measuring phenotype di↵erences between donors through molecular and omics assays. While villages show a lot of promise for uncovering Gene x Environment (GxE) interactions and the end e↵ect on disease, two main limitations need to be addressed: (i) experimental methods need to be refined to ensure reproducibility and adequate statistical power and (ii) statistical analysis tools need to be built to incorporate the full structure of these data. We aim to address these limitations by first developing standards for cell village experimental design through statistical modeling and simulation of village donor growth. Specifically, we will test how di↵erent experimental parameters a↵ect the amount of variation in donor growth rate estimates, the consistency of these predictions and whether experiments can maintain donor representation over the desired time frames. These factors will be important as we want to help researchers design experiments that minimize the variance in the growth rate estimates, while maintaining high genetic diversity throughout the entire experiment. Using this experimentally validated simulation framework, we will build a user- friendly tool that enables researchers to input their own preliminary data to project how di↵erent experimental design choices might impact the signals they are able to detect in the data and if donors might drop out or take over the population. In the second aim, we will build a hierarchical model of single cell RNA-seq (scRNA-seq) time series count data. The scRNA-seq pipeline will focus on two main applications: (i) a statistical tool that addresses the sparsity in single cell time series count data while accounting for natural variation in villages, and (ii) a tool that estimates the sequencing depth and village design needed to optimize statistical power. This work will help researchers conduct reproducible, high powered cell village experiments while fully leveraging the structure of the data. Improving cell village experimental and statistical methods will give researchers a powerful platform for investigating individual disease risk and potential for specific therapeutic interventions.