Biology subjects

Picault, S.

Publications and source records attributed to Picault, S..

6 recordsLinked to original sources

bioRxiv2023

Leveraging artificial intelligence and software engineering methods in epidemiology for the co-creation of decision-support tools based on mechanistic models

Epidemiological modelling is a key lever for infectious disease control and prevention on farms. It makes it possible to understand the spread of pathogens, but also to compare intervention scenarios even in counterfactual situations. However, the actual capability of decision makers to use mechanistic models to support timely interventions is limited. This study demonstrates how artificial intelligence (AI) techniques can make mechanistic epidemiological models more accessible to farmers and veterinarians, and how to transform such models into user-friendly decision-support tools (DST). By leveraging knowledge representation methods, such as the textual formalization of model components through a domain-specific language (DSL), the co-design of mechanistic models and decision-support tools becomes more efficient and collaborative. This facilitates the integration of explicit expert knowledge and practical insights into the modelling process. Furthermore, the utilization of AI and software engineering enables the automation of web application generation based on existing mechanistic models. This automation simplifies the development of DST, as tool designers can focus on identifying users needs and specifying expected features and meaningful presentations of outcomes, instead of wasting time in writing code to wrap models into web apps. To illustrate the practical application of this approach, we consider the example of Bovine Respiratory Disease (BRD), a tough challenge in fattening farms where young beef bulls often develop BRD shortly after being allocated into pens. BRD is a multi-factorial, multi-pathogen disease that is difficult to anticipate and control, often resulting in the massive use of antimicrobials to mitigate its impact on animal health, welfare, and economic losses. The decision-support tool developed from an existing mechanistic BRD model empowers users, including farmers and veterinarians, to customize scenarios based on their specific farm conditions. It enables them to anticipate the effects of various pathogens, compare the epidemiological and economic outcomes associated with different farming practices, and decide how to balance the reduction of disease impact and the reduction of antimicrobial usage (AMU). The generic method presented in this article illustrates the potential of artificial intelligence (AI) and software engineering methods to enhance the co-creation of decision-support tools based on mechanistic models in veterinary epidemiology. The corresponding pipeline is distributed as an open-source software. By leveraging these advancements, this research aims to bridge the gap between theoretical models and the practical usage of their outcomes on the field. HighlightsO_LIAI make mechanistic epidemiological models usable to support decisions on farms C_LIO_LITextual knowledge representation fosters co-design of mechanistic models and tools C_LIO_LIAI and software engineering automate web app generation to ease disease control C_LI

bioinformatics↗

bioRxiv2023

Pig herd management and infection transmission dynamics: a challenge for modellers.

The control of epidemics requires a thorough understanding of the complex interactions between pathogen transmission, disease impact, and population dynamics and management. Mechanistic epidemiological modelling is an effective way to address this issue, but handling highly structured and dynamic systems, remains challenging. We therefore developed a novel approach that combines Multi-Level Agent-Based Systems (MLABS) with spatial and temporal organization, allowing for a tuned representation of the transmission processes amongst the host population. We applied this method to model the spread of a PRRSv-like virus in pig farms, integrating the clinical consequences (conception and reproduction failures), in terms of animal husbandry practices. Results highlighted the importance to account for spatial and temporal structuring and herd management policies in epidemiological models. Indeed, disease-related abortions, inducing reassignments of sows in different batches, was shown to enhance the transmission process, favouring the persistence of the virus at the herd level. Supported by a declarative Domain-Specific Language (DSL), our approach provides flexible and powerful solutions to address the issues of on-farm epidemics and broader public health concerns. The present application, based on a simple Susceptible-Exposed-Infected-Recovered (SEIR) model, opens the way to the representation of more complex epidemiological systems, including more specific features such as maternally derived antibodies, vaccination, or dual infections, along with their respective clinical consequences on the management practices.

ecology↗

bioRxiv2021

The African swine fever modelling challenge: objectives, model description and synthetic data generation

African swine fever (ASF) is an emerging disease currently spreading at the interface between wild boar and pig farms in Europe and Asia. Current disease control regulations, which involve massive culling with significant economic and animal welfare costs, need to be improved. Modelling enables relevant control measures to be explored, but conducting the exercise during an epidemic is extremely difficult. Modelling challenges enhance modellers ability to timely advice policy makers, improve their readiness when facing emerging threats, and promote international collaborations. The ASF-Challenge, which ran between August 2020 and January 2021, was the first modelling challenge in animal health. In this paper, we describe the objectives and rules of the challenge. We then demonstrate the mechanistic multi-host model that was used to mimic as accurately as possible an ASF-like epidemic, provide a detailed explanation of the surveillance and intervention strategies that generated the synthetic data, and describe the different management strategies that were assessed by the competing modelling teams. We then outline the different technical steps of the challenge as well as its environment. Finally, we synthesize the lessons we learnt along the way to guide future modelling challenges in animal health. HighlightsO_LIThe 1st modelling challenge in animal health mimics ASF spread with synthetic data C_LIO_LIA mechanistic spatially-explicit stochastic model was developed to represent ASF spread and control C_LIO_LISynthetic data concerned date and location of detected cases in pigs and wild boar C_LIO_LIModelling ASF at the wildlife-livestock interface illustrates the reservoir role of wild fauna C_LI

ecology↗

bioRxiv2019

EMULSION: transparent and flexible multiscale stochastic models in epidemiology

Stochastic mechanistic models are essential to predictive epidemiology, to better explore biological assumptions and anticipate effects of control measures on pathogen spread. Their development is usually long and iterative, involving multi-disciplinary knowledge integration. However, knowledge often becomes lost in model code, hindering late changes in assumptions and often making models appearing as black boxes to end-users. We introduce here EMULSION, an artificial intelligence-based software intended to help modellers focus on model design rather than programming. EMULSION defines a domain-specific language to make all components of an epidemiological model (structure, processes, parameters...) explicit as a structured text file. This file is readable by scientists from other fields (epidemiologists, biologists, economists), who can contribute to validate or revise assumptions at any stage of model development. It is then automatically processed by EMULSION generic simulation engine, preventing any discrepancy between model description and implementation. The modelling language and simulation architecture both rely on the combination of advanced artificial intelligence methods (knowledge representation and multi-level agent-based simulation), allowing several modelling paradigms (from compartment- to individual-based models) at several scales (up to metapopulations). The flexibility of EMULSION and its capability to support iterative modelling are illustrated here through examples of progressive complexity, including late revisions of core model assumptions. EMULSION is also currently used to model the spread of several diseases in real pathosystems (zoonoses such as Q fever or the vector-borne Rift Valley fever; or bovine respiratory diseases, with a focus on detection and treatment protocols). EMULSION provides a command-line tool for checking models, producing model diagrams, running simulations, and plotting outputs. Implemented in Python 3, EMULSION runs on Linux, MacOS, and soon Windows. It is released under Apache-2.0 license. A comprehensive documentation with installation instructions, tutorial and examples is available from: https://sourcesup.renater.fr/emulsion-public.

epidemiology↗

bioRxiv2018

Environmental heterogeneity drives tsetse fly population dynamics

A spatially and temporally heterogeneous environment may lead to unexpected population dynamics, and knowledge still is needed on which of the local environment properties favour population maintenance at larger scale. As regards pathogen vectors, such as tsetse flies transmitting human and animal African trypanosomosis, such a knowledge is crucial for proposing relevant management strategy. We developed an original mechanistic spatio-temporal model of tsetse fly population dynamics, accounting for combined effects of spatial complexity, density-dependence, and temperature on the age-structured population, and parametrized with field and laboratory data. We confirmed the strong impact of temperature and adult mortality on tsetse populations. We showed that patches with the lowest mean temperatures and lowest variations act as refuges when adult mortality is homogeneously increased. Our results highlighted the importance of baseline data collection to characterize the targeted ecosystem before any control measure is implemented to maximize its efficiency.

ecology↗

bioRxiv2018

A generic multi-level stochastic modelling framework in computational epidemiology

There is currently an overwhelming increased interest in predictive biology and computational modelling. The development of reliable, reproducible and revisable simulation models in computational life sciences is often pointed out as a challenging issue. Population dynamics, including epidemiology, has not yet developed a language to formalize complex models in a univocal and automatable way, hence hindering the capability to implement in short time reliable, revisable and expert-friendly models intended for realistic mechanistic simulations. In epidemiology specifically, models aim not only at understanding pathogen spread but also at assessing control measures at several scales. To achieve this goal efficiently, best software practices should be supported by Artificial Intelligence methods to handle experts knowledge. The framework EMULSION presented here intends to both tackle multiple modelling paradigms in epidemiology and facilitate the automation of model design. We therefore built both a domain-specific language (DSL) for the modular description of complex epidemiological models, and a generic simulation engine designed to embed existing modelling paradigms within a homogeneous architecture based on adaptive software agents. The diversity of concerns (biology, economics, human activities) involved in real pathosystems requires an explicit, comprehensive and intelligible way to describe epidemiological models, to involve experts without computer science skills throughout the modelling, simulation and output analysis steps. This approach was applied to compare hypotheses in modelling a zoonosis (Q fever), to study its transmission dynamics within and between cattle herds at a regional scale, and to assess the contribution of transmission pathways. Separating model description from the simulation engine allowed epidemiologists to be involved in assumption revision, while guaranteeing very few code modifications. We assessed the added value of EMULSION by applying the DSL and the simulation engine to a concrete disease. Future extensions of EMULSION towards a broader range of epidemiological concerns will reduce significantly the time required to design and assess models and control measures against endemic and epidemic diseases. Ultimately, we believe this effort is a major lever to increase scientists preparedness to face emerging threats for public health and provide rapid, reliable, and reasoned assessments of control measures.

epidemiology↗