An Interpretable Graph-Regularized Optimal Transport Framework for Diagonal Single-Cell Integrative Analysis.

Key Points

• To evaluate the performance of GROTIA in integrating single-cell datasets using optimal transport and graph regularization techniques.
• Applied GROTIA to four simulated datasets and four real-world datasets
• Compared results with state-of-the-art unsupervised alignment methods
• Analyzed biological significance of top features in each dataset
• GROTIA outperformed existing unsupervised alignment methods
• Identified biologically relevant features in all tested datasets
• Confirmed the robustness of the alignment across different data types