Historically, human diseases have been described (and treated) based on their symptomology. Increasingly, however, elucidation of the molecular and functional underpinning of diseases is transforming our capacity to develop targeted therapeutic interventions. In the era of CRISPR, our ability to generate precise models of genetic diseases has been greatly improved, but unfortunately, a major bottleneck remains in our capacity to measure and interpret the resulting phenotypes. Moreover, most phenotypic measurements of animal models are qualitative, require expert human assessment, are narrowly focused based on the interests of individual labs, or are difficult to compare to other phenotypes. Together, these issues constrain our ability to generate the mechanistic genotype-phenotype maps that are necessary to understand how organisms are built, how health is maintained, and how these processes are disrupted in human diseases. Thanks to recent technological advances, it is now realistic to contemplate large-scale, systematic disease model generation and characterization. One of these advances is Multiplexed, Intermixed CRISPR Droplets (MIC-Drop), a technology that combines CRISPR, microfluidics, and barcoding to enable high-throughput gene disruption in zebrafish. By combining MIC-Drop with single-cell RNAseq and with automated physiological and behavioral assays, it is now possible to generate and characterize disease models with unprecedented depth and efficiency. Using this approach, we propose to target all human-disease-associated transcription factors in the zebrafish genome. This project will generate animal models for hundreds of human disease genes, along with single-cell transcriptomic data and detailed physiological and behavioral phenotypic descriptors for each model. These data will be analyzed using computational techniques to detect, categorize, and interpret mutant phenotypes that are otherwise intractable to manual curation. These data will be made available in a variety of formats to enable exploration of deep phenotypic data for all these disease models. We anticipate that this resource will improve the utilization, accessibility, and translational value of animal models to the research community.