Publications

PURPOSE OF REVIEW

Inflammatory breast cancer (IBC) is an uncommon but highly aggressive subtype of breast cancer that contributes significantly to breast cancer-related mortality. In this review, we provide an overview of the clinical and molecular characteristics of IBC, and highlight some areas of need for ongoing research.

RECENT FINDINGS

The disease is characterized by florid tumor emboli that obstruct dermal lymphatics, leading to swelling and inflammation of the affected breast. Recent studies have focused on tumor cell intrinsic features, such as signaling through pathways involved in growth and stem-like behavior, as well as extrinsic features, such as the immune system, that can be leveraged to develop new potential therapies. Key efforts have led to an increase in awareness of the disease as well as new insights into IBC pathogenesis. However, there is a strong need for new therapies designed specifically for IBC, and many unanswered questions remain.

Population dynamics can be inferred from genetic sequence data by using phylodynamic methods. These methods typically quantify the dynamics in unstructured populations or assume migration rates and effective population sizes to be constant through time in structured populations. When considering rates to vary through time in structured populations, the number of parameters to infer increases rapidly and the available data might not be sufficient to inform these. Additionally, it is often of interest to know what predicts these parameters rather than knowing the parameters themselves. Here, we introduce a method to  infer the predictors for time-varying migration rates and effective population sizes by using a generalized linear model (GLM) approach under the marginal approximation of the structured coalescent. Using simulations, we show that our approach is able to reliably infer the model parameters and its predictors from phylogenetic trees. Furthermore, when simulating trees under the structured coalescent, we show that our new approach outperforms the discrete trait GLM model. We then apply our framework to a previously described Ebola virus dataset, where we infer the parameters and its predictors from genome sequences while accounting for phylogenetic uncertainty. We infer weekly cases to be the strongest predictor for effective population size and geographic distance the strongest predictor for migration. This approach is implemented as part of the BEAST2 package MASCOT, which allows us to jointly infer population dynamics, i.e. the parameters and predictors, within structured populations, the phylogenetic tree, and evolutionary parameters.