PROJECT SUMMARY (See instructions): The emergence of single-cell RNA-sequencing (scRNA-seq) techniques has motivated many computational methods to study gene regulation and cell differentiation at the single-cell level. However, to improve the translational value of scRNA-seq, new methods are required to comparatively study the molecular differences between normal and pathological cells/tissues, and between control and case/treatment groups. The vast majority of existing network and clustering models have focused on scRNA-seq data generated under one experimental condition. Moreover, most existing scRNA-seq network models are correlative in nature and do not infer causality. What remains lacking are rigorous statistical methods for inference of differential causal gene regulation and cell composition in response to experimental interventions. There is, therefore, a critical need to develop novel methodologies for identifying the effects of experimental interventions on causal gene regulatory relationships and cell differentiation by jointly modeling scRNA-seq data across experimental groups. Without such tools, mechanistically understanding gene regulatory activities and cell differentiation will likely remain difficult. Our overall objective is to design and validate Bayesian network and clustering models for identifying differential causal gene regulatory networks and cell composition for scRNA-seq data generated under different experimental conditions. To achieve the overall objective, three specific aims will be pursued: Aim 1: Develop a differential zero-inflated negative binomial Bayesian network model to construct differential cell-type-specific causal gene regulatory networks under two experimental conditions. Aim 2: Develop a trajectory-dependent directed cyclic graph model to construct cell-specific causal networks with feedback loops and monitor its structural changes with cell development. Aim 3: Develop a scalable Bayesian semiparametric differential clustering method to discover differential cell composition and cell-type-specific marker genes that are shared and/or unique to each experimental group.