AIscEA: Unsupervised Integration of Single-cell Gene Expression and Chromatin Accessibility via Their Biological Consistency

Since the integrative analysis of single-cell gene expression and chromatin accessibility measurements is essential for revealing gene regulation at the single-cell resolution, integrating these two measurements becomes one of the key challenges in computational biology. Because gene expression and chromatin accessibility are measurements from different modalities, no common features can be directly used to guide their integration. Current state-of-the-art methods assume that the number of cell types across the measurements is the same. However, when cell-type heterogeneity exists, they might not generate reliable results. Furthermore, current methods do not have an effective way to select the hyper-parameter under the unsupervised setting. Therefore, applying computational methods to integrate single-cell gene expression and chromatin accessibility measurements remains difficult. We introduce AIscEA - Alignment-based Integration of single-cell gene Expression and chromatin Accessibility - a computational method that integrates single-cell gene expression and chromatin accessibility measurements using their biological consistency. AIscEA first defines a ranked similarity score to quantify the biological consistency between cell types across measurements. AIscEA then uses the ranked similarity score and a novel permutation test to identify the cell-type alignment across measurements. For the aligned cell types, AIscEA further utilizes graph alignment to align the cells across measurements. We compared AIscEA with the competing methods on several benchmark datasets and demonstrated that AIscEA is more robust to hyper-parameters and can better handle the cell-type heterogeneity problem. Furthermore, we demonstrate that AIscEA significantly outperforms the state-of-the-art methods when integrating real-world SNARE-seq and scMultiome-seq datasets in terms of integration accuracy.