Advances in high throughput omics technologies allow for assaying increasing compendium of molecular layers, from genome and epigenome profiling, transcriptomics to proteomics. Such data provide detailed snapshots which can characterize the molecular state for a given biology system from very fine resolution. Single cell genomics assays such as scRNA-seq and scATAC-seq specifically captures the landscape of genomic features across large collections of cells and have become one of the most popular molecular profiling techniques for investigating diverse problems related to gene regulation, such as identification of novel cell types and their regulatory signatures, trajectory inference for the analysis of continuous processes such as differentiation, high resolution analysis of transcriptional dynamics, and characterization of transcriptional heterogeneity within population of cells.

Despite the rapidly evolving technologies which can scales up to millions of cells across multiple individuals , one of the most pressing challenges in single cell genomics analysis is to address the amount of technical noise that can drive approximately 50% of the cell-cell variation in expression measurements. And such technical noise often times associated with high-sparsity of the genomic feature measurements. In chapter 2, we are mainly focusing on alleviating the effect of such technical variation in feature measurements of single cell genomics data, such as gene expression and locus accessibility. We show that this technical variation in both scRNA-seq and scATAC-seq datasets can be mitigated by analyzing feature detection patterns alone and ignoring feature quantification measurements. This result holds when datasets have low detection noise relative to quantification noise. We demonstrate state-of-the-art performance of detection pattern models using our new framework, scBFA, for both cell type identification and trajectory inference.

While single cell genomics assays are inherently high dimensional, the variations of individual cells are often summarized in a low dimensional space reflecting the change of gene’s mean expression. Gene co-expression networks, which often inferred from RNA sequencing data are another perspective to study cell type specific functional modules and complex regulatory interactions from transcriptomics profile. The increasing availability of large-scale scRNA-seq datasets is now making it possible to infer many gene networks from diverse cell populations. However, there are no mature tools currently available to visualize and compare large collections of networks across single cell populations, or for identifying correlations between variance in gene network structure with cell population-level phenotypes. In chapter 3, we present an unsupervised framework scMultiAE enabling comparison and visualization of multiple gene networks in a low-dimensional space with a focus on studying the heterogeneity of iPSCs during differentiation.

Single cell transcriptomic technologies which capture high dimensional measurements of gene expression in individual cells have been exponentially scaling in the number of cells that can be sequenced and analyzed simultaneously. Capturing a snapshot of the landscape for possible gene expression measurements from a collection of cells enables researchers to observe the space of molecular variation inherent to specific biological systems, termed atlasing. A challenge to building deeply characterized atlases of complex biological systems such as the human brain is in the identification and correction of confounding factors which do not relate to the underlying biology but instead arise from technical confounders. In this dissertation I present deep learning models applied to single cell genomics which remove unwanted technical variation and contamination as well as perform novel analysis not previously possible using standard methods. The construction of single cell genomics atlases leverages recent advances in single cell RNA sequencing technologies such as 10X and SmartSeq which can capture thousands of cells in single experiment. When the sequencing of individual cells is performed on different technologies this introduces unwanted technical variation (bias) specific to the technology and confounds attempts to merge scRNA-seq experiments into more complete atlases. To address this challenge, we developed scAlign to remove the effects of unwanted technical variation on gene expression specifically, scRNA-seq alignment based on advances in computer vision. scAlign, an unsupervised deep learning method, performs data alignment that can incorporate partial, overlapping or a complete set of cell labels, and estimate per-cell differences in gene expression across datasets or conditions to characterize specific expression changes due to conditions such as age or disease. With the recent surge of atlases efforts across complex tissues, conditions, and species another challenge is how to integrate the deep characterizations of cell state with lower resolution assays of single cell or bulk genomics. Specifically, spatial and multi-omics assays do not collect RNA from a single cell but instead from a spot containing multiple cells or in the later contamination from the unintended collection of additional cells. We developed scProjection to join deeply sequenced atlases with lower resolution genomic assays to address the unwanted heterogeneity in mixed samples and project such samples in a way that recovers the underlying single-cell measurements. scProjection is demonstrated to accurately estimate the abundance of cell types that compose a mixed RNA sample while simultaneously identifying the gene expression measurements consistent for each cell type in the sample to identify cell type specific changes due spatial location of cells or disease state.