Practical parameter identifiability and handling of censored data with Bayesian inference in mathematical tumour models.

Mechanistic mathematical models (MMs) are a powerful tool to help us understand and predict the dynamics of tumour growth under various conditions. In this work, we use 5 MMs with an increasing number of parameters to explore how certain (often overlooked) decisions in estimating parameters from data of experimental tumour growth affect the outcome of the analysis. In particular, we propose a framework for including tumour volume measurements that fall outside the upper and lower limits of detection, which are normally discarded. We demonstrate how excluding censored data results in an overestimation of the initial tumour volume and the MM-predicted tumour volumes prior to the first measurements, and an underestimation of the carrying capacity and the MM-predicted tumour volumes beyond the latest measurable time points. We show in which way the choice of prior for the MM parameters can impact the posterior distributions, and illustrate that reporting the most likely parameters and their 95% credible interval can lead to confusing or misleading interpretations. We hope this work will encourage others to carefully consider choices made in parameter estimation and to adopt the approaches we put forward herein.

Host resources and parasite traits interact to determine the optimal combination of host parasite-mitigation strategies.

Organisms have evolved diverse strategies to manage parasite infections. Broadly, hosts may avoid infection by altering behaviour, resist infection by targeting parasites or tolerate infection by repairing associated damage. The effectiveness of a strategy depends on interactions between, for example, resource availability, parasite traits (virulence, life-history) and the host itself (nutritional status, immunopathology). To understand how these factors shape host parasite-mitigation strategies, we developed a mathematical model of within-host, parasite-immune dynamics in the context of helminth infections. The model incorporated host nutrition and resource allocation to different mechanisms of immune response: larval parasite prevention; adult parasite clearance; damage repair (tolerance). We also considered a non-immune strategy: avoidance via anorexia, reducing intake of infective stages. Resources not allocated to immune processes promoted host condition, whereas harm due to parasites and immunopathology diminished it. Maximising condition (a proxy for fitness), we determined optimal host investment for each parasite-mitigation strategy, singly and combined, across different environmental resource levels and parasite trait values. Which strategy was optimal varied with scenario. Tolerance generally performed well, especially with high resources. Success of the different resistance strategies (larval prevention or adult clearance) tracked relative virulence of larval and adult parasites: slowly maturing, highly damaging larvae favoured prevention; rapidly maturing, less harmful larvae favoured clearance. Anorexia was viable only in the short term, due to reduced host nutrition. Combined strategies always outperformed any lone strategy: these were dominated by tolerance, with some investment in resistance. Choice of parasite mitigation strategy has profound consequences for hosts, impacting their condition, survival and reproductive success. We show that the efficacy of different strategies is highly dependent on timescale, parasite traits and resource availability. Models that integrate such factors can inform the collection and interpretation of empirical data, to understand how those drivers interact to shape host immune responses in natural systems.

No correlative evidence of costs of infection or immunity on leucocyte telomere length in a wild population of Soay sheep.

Telomere length (TL) is a biomarker hypothesized to capture evolutionarily and ecologically important physiological costs of reproduction, infection and immunity. Few studies have estimated the relationships among infection status, immunity, TL and fitness in natural systems. The hypothesis that short telomeres predict reduced survival because they reflect costly consequences of infection and immune investment remains largely untested. Using longitudinal data from a free-living Soay sheep population, we tested whether leucocyte TL was predicted by infection with nematode parasites and antibody levels against those parasites. Helminth parasite burdens were positively associated with leucocyte TL in both lambs and adults, which is not consistent with TL reflecting infection costs. We found no association between TL and helminth-specific IgG levels in either young or old individuals which suggests TL does not reflect costs of an activated immune response or immunosenescence. Furthermore, we found no support for TL acting as a mediator of trade-offs between infection, immunity and subsequent survival in the wild. Our results suggest that while variation in TL could reflect short-term variation in resource investment or environmental conditions, it does not capture costs of infection and immunity, nor does it behave like a marker of an individual's helminth-specific antibody immune response.

A unified framework for evolutionary genetic and physiological theories of aging.

Why and how we age are 2 intertwined questions that have fascinated scientists for many decades. However, attempts to answer these questions remain compartmentalized, preventing a comprehensive understanding of the aging process. We argue that the current lack of knowledge about the evolution of aging mechanisms is due to a lack of clarity regarding evolutionary theories of aging that explicitly involve physiological processes: the disposable soma theory (DST) and the developmental theory of aging (DTA). In this Essay, we propose a new hierarchical model linking genes to vital rates, enabling us to critically reevaluate the DST and DTA in terms of their relationship to evolutionary genetic theories of aging (mutation accumulation (MA) and antagonistic pleiotropy (AP)). We also demonstrate how these 2 theories can be incorporated in a unified hierarchical framework. The new framework will help to generate testable hypotheses of how the hallmarks of aging are shaped by natural selection.