Gut microbiota ecology: Biodiversity estimated from hybrid neutral-niche model increases with health status and aging

host. However, following the emerging evidence of its central role in human health and healthy aging [2–5], interest in the GM is considerably increased during the recent years and with it also the number of papers addressing its modelling, both through statistical and mechanistic approaches [6, 7]. While a straightforward theory of GM aging has not yet been designed, various efforts have been made to unveil how GM progresses with age [8] and to develop predictive models of the host biological age based on the GM taxonomic composition [9–11]. Interestingly, besides the specific taxonomic composition, also more global and structural properties of the GM, namely its biodiversity, have been associated with the host life-style, health status and aging [12, 13]. Biodiversity is a fundamental metric commonly used to model and quantify the health status of ecological systems [14, 15]. In ecosystems, in fact, biodiversity has a central role, safeguard- ing stability and resilience, ensuring sufficient variety of functional traits and species competi- tion, and preventing the predominance of invasive species [16–18]. It is hence not surprising that loss of microbial diversity is one of the most common GM unbalances in human diseases affecting westernized countries [12, 19], accompanied by alterations in GM stability and plas- ticity [2, 12]. Moreover, a connection between biodiversity and longevity has been hypothe- sized after the finding of an increase of subdominant species, including opportunistic and

allochthonous bacteria but also health-associated taxa, in longevity and extreme longevity [20]. Consistently, loss of GM biodiversity is usually associated with the overgrowth of pathogenic bacteria, fungi and other organisms that may favor excessive energy harvesting from ingested food and inflammatory response [12, 21]. Such disequilibrium has been linked to the emer- gence of the chronic and systemic low-grade inflammation named inflammaging, associated with morbidity and mortality in elderly people [4, 5, 22–24]. With the aim of investigating the relationship between the GM biodiversity and healthy aging, we consider here an ecological framework. In particular, trading off between the mini- mal degree of complexity and the maximal power of statistical prediction, we develop an eco- logical model that describes the GM population stationary distribution and allows to estimate its biodiversity. Notice that the biodiversity of an ecosystem can also be quantified using more classical approaches that do not rely on the assumption of an underlying dynamical process. There in fact exist various diversity indices that estimate biodiversity starting from the empirical relative abundances of species, such as Shannon [25], Simpson [26], Pielou [27] indices or the Hill’s numbers [28] (see Materials and Methods). However, a major advantage of the modeling approach is that biodiversity is computed directly from the inferred distribution rather than from the relative abundances. This method allows to mitigate the effect of data noise and vari-

ability (individual and populations) on the biodiversity index, thus increasing its robustness and general applicability. Here, we aim to identify a model that well describes the GM ecosystem, to assess whether the biodiversity estimate derived from such model is consistent with the most commonly used classical biodiversity indices and allows to achieve a higher statistical power. The modelling approach that we consider focuses on characterizing the GM empirical Rela- tive Species Abundance distribution (RSA), a curve that gained a central role in population modelling due to its similarity between different ecological systems. The RSA is constructed counting the number of species with a certain number of individuals and is often well described by long-tailed distributions belonging to the exponential family, such as the Log- Series or the Log-Normal [29]. To explain such regularity various statistical and dynamical models have been proposed [29, 30], and a particularly appealing approach is the one introduced by Volkov et al. [31], that relies on the neutrality [30] assumption. Following this hypothesis, the taxonomic differences PLOS ONE Gut microbiota biodiversity and healthy aging PLOS ONE | https://doi.org/10.1371/journal.pone.0237207 October 30, 2020 2 / 23 Funding: GC and DR were supported by the Interomics CNR Flagship Initiative PB05 (https:// www.interomics.eu/), by the European Union’s Seventh Framework Programme (grant agreement 305280 MIMOmics, http://www.mimomics.eu/) and by the European Union’s H2020 Programme (ITN grant agreement 721815 ImforFUTURE,

https://imforfuture.eu/). CF was supported by the grant of the Ministry of Education and Science of the Russian Federation Agreement No. 075-15- 2019-871 (http://government.ru/en/). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interests: The authors have declared that no competing interests exist.