Summary Advanced single-cell sequencing techniques have enabled us to infer gene regulation at the single-cell level. We propose to develop computational methods to overcome obstacles for elucidating gene regulation at single-cell resolution. We first present an alignment-based computational framework to integrate single-cell multi-omics measurements. The alignment-based computational framework can effectively handle the cell type imbalance problem and is more robust to hyperparameters. Furthermore, we incorporate the integrated single-cell multi-omics measurements and advanced machine learning algorithms to infer transcriptional regulation, distal regulatory elements, and post-transcriptional regulation at the single-cell level. We expect to develop computational methods to better understand gene regulation, which would lay a solid foundation for disease diagnosis, treatment, and prevention.