The project, funded by the German Federal Ministry of Education and Research (BMBF) under grant 031L0170C, ran from 1 April 2019 to 31 December 2022 at the Technical University of Munich’s Institute for Medical Microbiology, Immunology and Hygiene. It was part of the e:Bio – Computational Life Science initiative, specifically the TIDY toolbox for inference of cellular dynamics in tissues (subproject C). The research team included experts from the university and collaborators such as Prof. Thomas Höfer, Dr. Carsten Marr, Prof. Dirk Busch, Dr. Veit Buchholz and Prof. Markus Gerhard, who contributed experimental data and methodological insights.

The core scientific aim was to develop mathematical tools that can extract in‑vivo cell‑proliferation rates from population‑level experiments. Two main labeling strategies were addressed. First, dual DNA labeling with the thymidine analog BrdU and the DNA‑binding dye 7‑Aminoactinomycin allowed the authors to derive a compact formula for the cell‑cycle length: (p approx 2lambdaDelta t – 2p_{text{G2M}}), where (p) is the fraction of DNA(^{2N})/BrdU(^+) cells, (lambda) the proliferation rate, (Delta t) the interval between BrdU injection and DNA measurement, and (p_{text{G2M}}) the fraction of cells in G2/M. A formalism was then built to calculate expected fractions in each DNA/BrdU quadrant, enabling the estimation of mean durations of all cell‑cycle phases from a single time‑point measurement. Monte‑Carlo simulations validated this approach, confirming its accuracy and allowing optimisation of experimental parameters. The method proved robust even when cell‑death times followed strongly non‑exponential, log‑normal distributions, with errors remaining below 4 % under realistic coefficient‑of‑variation regimes.

Second, the project tackled generation‑resolved labeling with CFSE and Cell Trace Violet (CTV). Flow‑cytometry data from these dyes contain overlapping fluorescence peaks corresponding to successive cell divisions. Advanced deconvolution algorithms were developed to account for population heterogeneity and to separate individual generations. The resulting proliferation rates were cross‑validated against the dual‑label DNA method.

A key biological application involved T‑cell dynamics during acute infection. OT‑I T cells were transferred into C57BL/6 mice and infected with Listeria monocytogenes expressing ovalbumin. BrdU was administered at defined times before analysis, and flow cytometry provided DNA/BrdU profiles. The new framework quantified division times of approximately 8.55 ± 0.95 h, with BrdU availability of one hour and analysis at three hours post‑labeling. By incorporating phospho‑Rb measurements to identify resting cells, the authors corrected for quiescence and revealed a progressive slowdown of both central memory (CMP) and non‑central memory (NCMP) T‑cell subsets over days 4, 5, and 8 post‑infection. In a parallel experiment, central memory T cells were stimulated under normal conditions or with diphtheria toxin‑mediated suppression of antigenic stimulus; the method captured the resulting differences in proliferation rates.

Overall, the project delivered a suite of ODE‑based parameter‑estimation tools that translate complex flow‑cytometry data into precise, time‑resolved proliferation metrics. These tools, validated through extensive simulation and applied to in‑vivo immunological models, provide a powerful framework for studying cell‑cycle dynamics in health and disease. The collaboration among the university’s core team and external experts ensured rigorous experimental design, data acquisition, and methodological development, culminating in a robust, publicly available toolbox for the broader scientific community.