Manifold learning analysis suggests novel strategies for aligning single-cell multi-modalities and revealing functional genomics for neuronal electrophysiology

Abstract

Recent single-cell multi-modal data reveal multi-scale characteristics of single cells, such as transcriptomics, morphology, and electrophysiology. However, our understanding of functional genomics and gene regulation leading to various cellular characteristics remains elusive. To address this, we applied multiple machine learning methods to align gene expression and electrophysiological data of single neuronal cells in the mouse brain. We found that nonlinear manifold learning outperforms other methods. After manifold alignment, the cell clusters highly correspond to transcriptomic and morphological cell-types, suggesting a strong nonlinear relationship between gene expression and electrophysiology at the cell-type level. The aligned cells form developmental trajectories and show continuous changes of electrophysiological features, implying the underlying developmental process. We also found that the manifold-aligned cell clusters differentially expressed genes can predict many electrophysiological features. Functional enrichment and gene regulatory network analyses for those cell clusters revealed potential genome functions and molecular mechanisms from gene expression to neuronal electrophysiology.