770-Genes2Genes: for Single-Cell Trajectory Alignment

The paper introduce Genes2Genes (G2G), a novel Bayesian information-theoretic framework designed to align and compare single-cell transcriptomic trajectories. Unlike traditional methods like dynamic time warping, which struggle with mismatched data, G2G uses a five-state dynamic programming algorithm to precisely identify matches, warps, and gaps such as insertions or deletions. This approach allows researchers to pinpoint exactly where in vitro cell engineering diverges from in vivo development or how disease states differ from healthy controls. By analyzing individual gene expression distributions rather than just mean values, the tool captures sequential matches and mismatches at a single-gene resolution. The framework also includes features for clustering similar alignment patterns and performing downstream biological pathway analysis. Ultimately, G2G outperforms existing state-of-the-art tools by providing more accurate, statistically consistent comparisons of dynamic cellular processes.References: Sumanaweera D, Suo C, Cujba A M, et al. Gene-level alignment of single-cell trajectories[J]. Nature Methods, 2025, 22(1): 68-81.前往小宇宙评论区与主播互动