Poster #79 - Emilia (Xinyi) Chen

Resolution of Selection Versus Adaptation in Cellular Evolution

Xinyi E. Chen1,6,7,†, Kevin Z. Lin2,†, Dylan Schaff3,4†, Robert Vander Velde5, Christopher Cote5, Sijia Huang6, Andy J. Minn7,8,9,10, Sydney M. Shaffer3,5,∗, Nancy R. Zhang8,10,11, * 1Graduate Program in Genomics and Computational Biology, University of Pennsylvania 2Department of Biostatistics, University of Washington 3Department of Bioengineering, University of Pennsylvania 4Merck & Co, Incorporated 5Department of Pathology and Laboratory Medicine, University of Pennsylvania 6Institute for Biomedical Informatics, University of Pennsylvania 7Department of Radiation Oncology, University of Pennsylvania 8 Abramson Family Cancer Research Institute, University of Pennsylvania 9 Parker Institute for Cancer Immunotherapy, University of Pennsylvania 10 Mark Foundation Center for Immunotherapy, Immune Signaling, and Radiation, University of Pennsylvania 11Department of Statistics and Data Science, University of Pennsylvania †These authors contributed equally to this work. ∗Co-correspondence

Cells exposed to drugs or other external stress evolve through population bottlenecks where a small minority can determine the fate of the whole population. It is difficult to tell whether survivors prevail by changing state (adaptation) or by the selective outgrowth of cells already poised to persist. Yet, dissecting the relative contributions of selection versus adaptation is fundamental to understanding cellular evolution and can directly impact treatment design. Here, we introduce a lineage-resolved single-cell framework that estimates each cell's contribution to a future population and quantifies the extent to which selection and adaptation shape that outcome. Applied to a cancer resistance model, we identify clones primed to endure treatment versus ones that survive by producing diverse progeny. We further show that early survivors and long-term persisters often arise from different clones with distinct molecular programs, that the relative weight of selection and adaptation varies by treatment and by clone, and that some stress-response pathways reverse their roles between short- and long-term treatment. Analyses of patient cohorts show the same time-dependent role reversal of genetic programs. This contrast between selection and adaptation, and their distinct genetic programs, is clinically relevant, potentially guiding which features to target in combination and which to disrupt through treatment sequencing.