Abstract 7500: Joint reconstruction of single-cell evolutionary lineages and fate transition trajectories from long-read single cell transcriptomes

Abstract Computational inference of endogenous genetic alterations from single cell transcriptomic data opens an avenue of measuring both evolution and fate markers in individual cells, in particular, the emerging long-read single cell RNA sequencing (LR-scRNAseq) technologies allowed detection of multi-layered mutations across whole transcript bodies of thousands of genes in thousands of cells in parallel. However, robust tools for integrating single cell genetic evolutionary lineage and fate transition trajectory from these powerful data are missing. To address this, we developed EvoFATE, an unsupervised graph learning model, to automatically reconstruct unified roadmaps of single cell evolutionary lineages and fate transition trajectories by discovering representations using endogenous mutations as evolutionary markers and gene expressions as fate remodeling markers. EvoFATE recovered complex evolutionary fate dynamics in simulations and blood samples where diverse cell types had known lineage orders. We applied EvoFATE onto LR-scRNAseq data of patient-matched lung metastatic tumors and identified the evolutionary fate paths of tumor cells during both metastatic seeding and clonal expansion stages. To summarize, EvoFATE robustly integrates genetic evolutionary lineages with cell fate transition trajectories to resolve complex fate remodeling dynamics and evolution-fate synchronies using cell endogenous markers, enabling precise tracing of temporal changes in cells under real pathophysiology conditions without genetic manipulations. Citation Format: Yueying He, Yuanqing Yan, Cheng-Kai Shiau, Lina Lu, Minhua Wang, Taisuke Kaihou, Chitaru Kurihara, Kazutaka Fukumura, Rajul K. Ranka, Yudai Miyashita, Timothy Pan, Arvind Bhimaraj, Daniel Brat, Jason Huse, Ankit Bharat, Ruli Gao. Joint reconstruction of single-cell evolutionary lineages and fate transition trajectories from long-read single cell transcriptomes [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 7500.