Fig. 3: scGALA Serves as a Universal Booster for Existing Single-Cell Analysis Pipelines by Replacing Cell Alignment Modules.

a UMAPs comparing batch correction by INSCT (supervised baseline) and scGALA-enhanced INSCT (replacing alignment module), colored by batch (left) and cell type (right). b Normalized metrics (0-1) quantifying biological conservation (e.g., cell type purity) and batch effect removal. Higher values indicate better performance. c Label transfer accuracy (Spearman correlation vs. ground truth) comparing Seurat (baseline) and scGALA-enhanced Seurat, with UMAPs colored by transferred cell type labels. d Label transfer performance metrics: accuracy, Cohen’s kappa, and 95% confidence intervals (box plots). e Multi-omics integration (RNA + ATAC) comparing Seurat (baseline) and scGALA-enhanced Seurat, colored by modality domain (left) and joint cell type labels (right). f Multi-omics integration metrics: biological conservation and domain harmonization (box plots), and FOSCTTM (lower is better; line plot) across varying cell numbers. g Spatial alignment of tissue slices: spatial coordinate plots (X/Y/Z tissue positions) comparing STAligner (baseline) and scGALA-enhanced STAligner, colored by spatial region labels. h Spatial integration metrics: biological conservation, modality mixing, and alignment accuracy (box plots). Each score in panels b, d, f, and h is derived from N = 10 bootstrapping replicates using different random seeds (technical replicates). Boxes indicate the interquartile range (IQR, 25th to 75th percentile), with the line inside each box representing the median. Whiskers extend to the most extreme data points within 1.5 times the IQR from the quartiles. P values calculated using one-sided Student’s t-test: * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001; ns, not significant. Exact P values are provided in the Source Data. Source data are provided as a Source Data file.