The recent advances in single-cell sequencing technologies provide
unprecedented opportunities to decipher the multi-scale gene regulatory
grammars at diverse cellular states. Here, we will introduce our computational
efforts to decipher cell-type-specific gene regulatory grammar using large-scale
single-cell multi-omics data. First, we developed a deep generative model,
named SAILER, to learn the low-dimensional latent cell representations from
single-cell epigenetic data for accurate cell state characterization. SAILER
adopted the conventional encoder-decoder framework and imposed additional
constraints for biologically robust cell embeddings invariant to confounding
factors. Then, we will introduce DIRECT-NET, an efficient method to discover
cis-regulatory elements and construct regulatory networks using single-cell
multi-omics data. Unlike existing methods requiring extensive functional
genomic data, DIRECT-NET can build cell-type-specific gene regulatory
networks from individual genomes without any auxiliary data. Finally, we
applied our methods on single cell data and discovered key genetic and
epigenetic changes in brain disorders.