eGUT: a Tool for Predictive Computer Simulation of the Gut Microbiota and Host Interactions
Lead Research Organisation:
University of Birmingham
Department Name: Sch of Biosciences
Abstract
We want to build, apply, and disseminate eGUT, a new computer simulation platform for the animal and human gut, to reduce and replace animal experiments.
The gut, in particular the large intestine, is teeming with microorganisms, collectively called the microbiota: 50% of faeces are microorganisms. Overall the indigenous microorganisms called commensals are beneficial for the host, illustrated by the fact that germ-free animals develop rather poorly. It is the behaviour of the microbiota and their metabolic and other effects on the host which we want to simulate in eGUT. The host can be represented by the gut wall (the mucosa), as all interactions are mediated through the gut wall.
Due to an increasing awareness of the importance of the gut microbiota, the number of studies using animal experiments has risen sharply, so that the need to develop alternatives to animal experiments is becoming increasingly urgent.
It is very difficult to estimate the number of animal experiments that eGUT will avoid, so the numbers should be treated with caution. We estimate that eGUT, once fully developed and accepted, could save up to 150,000 animals globally per year. However, it will take some thorough case studies to improve and validate eGUT and to demonstrate that eGUT can make reliable predictions of animal experiments based on in vitro laboratory data, before the use of eGUT will become widely accepted. We will therefore focus on these case studies to make a convincing case for the use of eGUT. After this, more and more scientists will adopt it, mainly to reduce the numbers of animals used, but eventually to fully replace a range of animal experiments, for example in the area of nutrition.
Objectives:
1. Build eGUT
eGUT will be based on the long experience of JK and collaborators with developing individual-based models of the interactions of microorganisms. It will build on our state of the art individual-based Dynamics of Microbial Communities Simulator iDynoMiCS, which was published in 2011 and released for public use (open source) on www.idynomics.org. While iDynoMiCS is an excellent basis for eGUT, a number of additions to simulate the gut will be necessary (add a description of the gut wall, add superindividuals and related procedures to allow simulation of a large system, couple several simulations of individual gut compartments into a multi-compartment simulation). Moreover, we want eGUT to be more user-friendly by adding a graphical user interface and by providing more detailed documentation and hands-on tutorials. This is time consuming but very important to allow scientists without mathematical or computer science backgrounds to use eGUT, and this functionality will enable a far wider number or users and therefore far more impact on the 3Rs.
2. Apply eGUT
Because we need test cases for eGUT to validate it and to convince others that it makes reliable predictions, we will apply eGUT to three test cases. More than one is necessary to show that it does not just work for one specific problem, but it is important to not attempt too many as each case will require time if it is to be done well. The test cases proposed are: one from our own laboratory; one from Prof Hauke Smidt at Wageningen University in The Netherlands; and one from Prof Tine Licht from the National Food Institute at the Technical University of Denmark.
3. Disseminate eGUT
Dissemination is very important to build up a user base for eGUT, so we will run two one-week workshops where we will teach potential users of eGUT how they can apply it to their own research in order to reduce or replace animal experiments. Further, we will build a dedicated web site and use media channels to promote eGUT (blogs, YouTube, popular science magazines), amongst other plans for dissemination.
The gut, in particular the large intestine, is teeming with microorganisms, collectively called the microbiota: 50% of faeces are microorganisms. Overall the indigenous microorganisms called commensals are beneficial for the host, illustrated by the fact that germ-free animals develop rather poorly. It is the behaviour of the microbiota and their metabolic and other effects on the host which we want to simulate in eGUT. The host can be represented by the gut wall (the mucosa), as all interactions are mediated through the gut wall.
Due to an increasing awareness of the importance of the gut microbiota, the number of studies using animal experiments has risen sharply, so that the need to develop alternatives to animal experiments is becoming increasingly urgent.
It is very difficult to estimate the number of animal experiments that eGUT will avoid, so the numbers should be treated with caution. We estimate that eGUT, once fully developed and accepted, could save up to 150,000 animals globally per year. However, it will take some thorough case studies to improve and validate eGUT and to demonstrate that eGUT can make reliable predictions of animal experiments based on in vitro laboratory data, before the use of eGUT will become widely accepted. We will therefore focus on these case studies to make a convincing case for the use of eGUT. After this, more and more scientists will adopt it, mainly to reduce the numbers of animals used, but eventually to fully replace a range of animal experiments, for example in the area of nutrition.
Objectives:
1. Build eGUT
eGUT will be based on the long experience of JK and collaborators with developing individual-based models of the interactions of microorganisms. It will build on our state of the art individual-based Dynamics of Microbial Communities Simulator iDynoMiCS, which was published in 2011 and released for public use (open source) on www.idynomics.org. While iDynoMiCS is an excellent basis for eGUT, a number of additions to simulate the gut will be necessary (add a description of the gut wall, add superindividuals and related procedures to allow simulation of a large system, couple several simulations of individual gut compartments into a multi-compartment simulation). Moreover, we want eGUT to be more user-friendly by adding a graphical user interface and by providing more detailed documentation and hands-on tutorials. This is time consuming but very important to allow scientists without mathematical or computer science backgrounds to use eGUT, and this functionality will enable a far wider number or users and therefore far more impact on the 3Rs.
2. Apply eGUT
Because we need test cases for eGUT to validate it and to convince others that it makes reliable predictions, we will apply eGUT to three test cases. More than one is necessary to show that it does not just work for one specific problem, but it is important to not attempt too many as each case will require time if it is to be done well. The test cases proposed are: one from our own laboratory; one from Prof Hauke Smidt at Wageningen University in The Netherlands; and one from Prof Tine Licht from the National Food Institute at the Technical University of Denmark.
3. Disseminate eGUT
Dissemination is very important to build up a user base for eGUT, so we will run two one-week workshops where we will teach potential users of eGUT how they can apply it to their own research in order to reduce or replace animal experiments. Further, we will build a dedicated web site and use media channels to promote eGUT (blogs, YouTube, popular science magazines), amongst other plans for dissemination.
Technical Summary
We propose the development of a generic computer simulation tool to reduce and replace animal experiments by simulating the gut microbiota and their interactions with the host. This tool, called eGUT, will be built by extending the individual-based Dynamics of Microbial Communities Simulator iDynoMiCS, which we have developed with many colleagues and published in Environmental Microbiology last year and made available as open source on www.idynomics.org. Using an individual-based model (IbM) as the engine for the tool has the advantage that the tool will be very flexible, e.g., the number of species or the types of interactions or the rules of behaviour can be chosen by the scientist user - as required of a generic tool. Moreover, IbMs merely describe the activities of the individual organisms or host cells. Their behaviour is described by equations for the kinetics of growth and by rules for cell division (if larger than threshold divide) and other decisions. These kinetics and rules are typically based on laboratory experiments under defined conditions. It is quite important to realize that simple rules can lead to complex dynamics and spatial patterns so that complexity of a system does not mean that it cannot be understood or modelled because it is too complex. Complexity simply emerges from the multitude of local interactions of various individuals with the environment and with the host cells, as running an IbM will demonstrate.
The gut microbiota has profound effects on our health, a realization leading to a surge of animal experiments in this area, so that the development of a computer simulation tool to reduce or replace these animal experiments is a gap that is becoming increasingly urgent to fill.
Our specific aims are to build eGUT, apply eGUT to model the spread and replacement of antibiotic resistance plasmids in the gut, apply eGUT to two further test cases from our collaborators, and to disseminate eGUT to actively increase the user base.
The gut microbiota has profound effects on our health, a realization leading to a surge of animal experiments in this area, so that the development of a computer simulation tool to reduce or replace these animal experiments is a gap that is becoming increasingly urgent to fill.
Our specific aims are to build eGUT, apply eGUT to model the spread and replacement of antibiotic resistance plasmids in the gut, apply eGUT to two further test cases from our collaborators, and to disseminate eGUT to actively increase the user base.
Planned Impact
This project will initially help to reduce and then increasingly help to replace animal experiments with computer simulations.
The scale of reduction is difficult to estimate, but for the UK we estimate the potential reduction and replacement to be about 11,000 animals per year, and globally the potential reduction could be about 150,000 per year. This assumes that experiments in the area of gut microbiota host interaction do not continue to rise as they currently do, which is very conservative. On the other hand, the estimate assumes that eGUT is used whenever possible, which is very optimistic as currently the acceptance of computer simulations is not very good. We believe that acceptance and use of computer simulations will grow over time with every successful computer simulation case study, so we think the figure can be reached in the long term. The beginning will be rather more modest, but the sooner one starts this positive feedback loop the better.
For nutrition studies that we think can be mostly replaced by eGUT, mainly chicken, mice, rats, and fish are used, but in lower numbers also a range of other animals, e.g. pigs and ruminants.
Locally, our own laboratory is currently building in vitro models of the gut to study the spread of plasmids we are developing, with the help of Wellcome Trust ISSF funding, to replace other plasmids that carry antibiotic resistence genes and that are prevalent in farm animals as well as humans in various countries. Getting rid of resistances is an alternative to finding new antibiotics, which is getting increasingly difficult. After evaluating and optimizing the technology in in vitro models, we will finally need to show that it also works in real animals, which will be the subject of a future grant application to BBSRC. This need prompted us to consider developing eGUT in the first place because it will enable us to reduce the number of pigs we require to the minimum.
The scale of reduction is difficult to estimate, but for the UK we estimate the potential reduction and replacement to be about 11,000 animals per year, and globally the potential reduction could be about 150,000 per year. This assumes that experiments in the area of gut microbiota host interaction do not continue to rise as they currently do, which is very conservative. On the other hand, the estimate assumes that eGUT is used whenever possible, which is very optimistic as currently the acceptance of computer simulations is not very good. We believe that acceptance and use of computer simulations will grow over time with every successful computer simulation case study, so we think the figure can be reached in the long term. The beginning will be rather more modest, but the sooner one starts this positive feedback loop the better.
For nutrition studies that we think can be mostly replaced by eGUT, mainly chicken, mice, rats, and fish are used, but in lower numbers also a range of other animals, e.g. pigs and ruminants.
Locally, our own laboratory is currently building in vitro models of the gut to study the spread of plasmids we are developing, with the help of Wellcome Trust ISSF funding, to replace other plasmids that carry antibiotic resistence genes and that are prevalent in farm animals as well as humans in various countries. Getting rid of resistances is an alternative to finding new antibiotics, which is getting increasingly difficult. After evaluating and optimizing the technology in in vitro models, we will finally need to show that it also works in real animals, which will be the subject of a future grant application to BBSRC. This need prompted us to consider developing eGUT in the first place because it will enable us to reduce the number of pigs we require to the minimum.
