Research in the Hart Lab has focused on the central concepts of modular cell biology, as put forward by Hartwell et al, 1999: how normal cells rely on an interconnected web of biological processes for survival and proliferation, and how mutation rewires this web of dependencies into disease states. We have developed experimental and computational tools for CRISPR- mediated perturbation studies that provide an understanding of the hierarchical organization of the mammalian cell and reveal context-specific genetic vulnerabilities. Our work can be reasonably divided into research on first-order effects of gene perturbation, accurately measuring gene essentiality and differential essentiality, and second- order effects including digenic interaction, functional buffering, and network approaches. We developed the TKOv3 CRISPR/Cas9 genome-scale library for knockout screens in human and mouse cells (Hart et al, 2017), as well as the BAGEL (Kim & Hart, 2021) and DrugZ (Colic et al, 2019) software packages for analysis of fitness and chemogenetic interaction screens. Our integrative analysis of hundreds of cell-line screens from the Cancer Dependency Map initiative yielded one of the first coessentiality maps describing functional linkages between human genes (Kim et al, 2019); the first systematic survey of proliferation suppressor genes, discovering a novel putative tumor suppressor role for saturated fatty acid synthesis in myeloid leukemia (Lenoir et al, 2021); and one of the first integrated computational and experimental studies confirming that functional buffering by, e.g., paralogs is systematically missed by monogenic CRISPR/Cas9 knockout screens (Dede et al, 2020). These latter two works used the Cas12a CRISPR endonuclease and its endogenous multiplexing capability to build efficient assays for genetic interaction between targeted gene pairs. Our future work will deepen our understanding of how both first-order and second-order effects shape modular biology, and improve our ability to decipher the natural complexity of the cell by extending this work into higher-order effects from targeted polygenic perturbations. First- generation network approaches integrate functional linkage across all contexts; future computational work will decipher lineage-specific interactions to define more precise cellular networks for functional genomics and tissue-specific disease modeling. On the experimental side, we will continue to push the state of the art in genetic perturbation technology, developing a highly multiplexed and multimodal perturbation platform that can go beyond digenic interactions and provide deeper insight into the complexity of the mammalian cell.