Theoretical Systems Biology

Present research priorities
Given that cancer-associated mutations occur continuously in renewing tissues, such as blood, these tissues are likely to have protective mechanisms in place that limit the selection of cancer drivers and malignant transformation. To understand the nature of this protection, we study the dynamics of non-malignant, steady-state and stress, hematopoiesis in mice and humans, for both cell clones (Busch et al., 2015; Pei et al., 2017) and the underlying molecular regulation. Moreover, using mouse models of sporadic leukemia and human blood and bone marrow samples, we investigate how perturbations in normal stem and progenitor cell dynamics instigate the development of acute leukemia. Methods to study somatic evolution in humans, using population genetics of somatic mutations, carry over to solid tissues, allowing us to study the evolution of cancer in the developing and adult nervous systems (Körber et al., 2023). Our research on clonal dynamics in somatic evolution was initiated by research on T cell responses at single-clone resolution. Here, we devised a mathematical inference framework predicting from experimental data that memory T cells begin to develop early during an acute immune response, from a stem-like T cell subset (Buchholz et al. 2013). This view has now gained broad acceptance, and we continue to dissect the clonal dynamics of adaptive immune responses and their regulation. All our projects involve close collaboration with experimentalists and theoreticians.  

Current developments
The computational reconstruction of somatic evolution from genome sequencing data allows us the timing of key genetic events in oncogenesis. In collaboration with researchers at the Hopp Children’s Cancer Center, we are currently exploring the use of these methods for early risk prediction for pediatric cancers. Furthermore, we are developing robust methods to detect clonal selection in normal tissues.  

Methods and technologies
We make use of a broad array of methods for the mathematical modeling of clonal dynamics and somatic evolution and for confronting these models with experimental data, using Bayesian inference and artificial intelligence. For in situ barcoding studies in mice, we developed a suite of computational tools, tailored to the Polylox and PolyloxExpress systems (Pei et al., 2017).

Goals and societal relevance
We expect that understanding the mechanisms that allow pre-malignant clones to grow in normal tissues and eventually develop into full-blown malignant tumors will allow us to develop new methods for the earlier detection of cancer, in turn enabling earlier and more effective treatment.