TyCHE enables time-resolved lineage tracing of B cells in primary and recall germinal center reactions 2261047

Abstract Introduction Phylogenetic methods for cell lineage tracing have driven significant insights into B cell development, immune responses, and viral evolution. They have been used to identify the ignition date of the HIV-1 pandemic, and the spread of SARS-CoV-2 globally. While most methods estimate mutation trees, time-resolved lineage trees are more interpretable and could relate cellular migration and differentiation to perturbations like vaccines and drug treatments. However, somatic mutation rates vary dramatically by cell type, significantly biasing existing methods. For example, B cells undergo periods of rapid somatic hypermutation during immune responses before becoming quiescent memory cells. Methods We introduce TyCHE (Type-linked Clocks for Heterogenous Evolution), a Bayesian phylogenetics package that infers time-resolved phylogenetic trees of populations with distinct evolutionary rates. To validate our model, we developed an agent-based simulation package, SimBLE. Results Using SimBLE simulations, we show TyCHE estimates more accurate tree topologies, node dates, and ancestral cell types than existing methods. Further, we show how TyCHE can use BCR sequences to accurately reconstruct both primary GC reactions in HIV infection and recall GC reactions following influenza vaccination. We also use TyCHE to infer clinically realistic temporal evolution of a glioma tumor lineage and progression of a bacterial lung infection. Conclusion TyCHE is tailored to the unique challenges of B cell phylogenetic inference and generates highly accurate trees for B cell and non-B cell heterogeneously evolving populations. These advances in time-resolved phylogenetic inference open new avenues for understanding how cell proliferation, differentiation, and migration shape responses to vaccines, drug treatments, and other stimuli. TyCHE and SimBLE are available as open-source software packages compatible with the BEAST2 and Immcantation ecosystems. Funding Source National Institute for Allergy and Infectious Diseases grant R00AI159302 Topic Categories Computational and Systems Immunology (COMP)