The open-source BioDynaMo platform was developed at CERN to assist life scientists in creating biological simulations. Since the start of the COVID-19 pandemic, the platform has been adapted to simulate how the coronavirus spreads in populations, which can help to control the pandemic and inform decisions for similar outbreaks in the future.
- In life sciences, the ‘single researcher’s project’ approach may not be efficient for the development of biomedical research.
- BioDynaMo’s open-source design facilitates the examination of biological models by minimising the coding efforts for researchers.
- Against the pandemic background, BioDynaMo is being applied to simulate various epidemic scenarios using the SEIR model.
- A simulation of SARS-CoV-2 spread in closed environments is presented.
BioDynaMo started as a collaborative knowledge transfer project at CERN, with the goal to ‘share knowledge’ that is present at CERN in the areas of computer simulations, efficient and scalable software development and long-running sustainable software collaborations with the fields of life science. The main problem at hand was the absence of a standardised and high-performance platform for conducting in-silico biomedical experiments (i.e. simulations). Simulation is an indispensable tool in aiding biomedical researchers to understand complex biological systems and, ultimately, to develop new medicine. Life scientists traditionally follow the ‘single researcher’s project’ approach, in which a model is developed to investigate a specific scientific question and is abandoned after the question has been answered and the work has been published. This inhibits other scientists from building upon prior work and effectively slows down the pace of biomedical research, making it a societal problem at large.
BioDynaMo is an open-source C++ framework where life scientists can easily create, run and visualise 3D agent-based biological simulations. It was designed so that users can examine their biological models with minimal coding effort and rely on our highly optimised execution engine that deals with the intricacies involved in the world of high-performance computing. The compute-intensive part of mechanical interactions in the BioDynaMo code base has already been made compatible to run on graphics processing units (GPUs). In order to push the boundaries of biomedical research even further, we are now working on accelerating extracellular diffusion computations on GPUs. An example of a mechanism addressed by BioDynaMo is that of predicting the growth and the 3D morphology of a tumour as shown in Figure 1.
Our platform enables the simulation of 1.24 billion agents on a single server and 12 million agents on a laptop. BioDynaMo places a lot of focus on hiding computational complexity and providing an easy-to-use interface, such that the life scientist can concentrate on biological aspects, rather than computational. BioDynaMo helps scientists translate an idea quickly into a simulation by providing common building blocks, and a modular and extensible software design. An analysis of the performance of the platform and demonstration of its capabilities with three example use cases: soma clustering, neural development, and tumour spheroid growth, is presented in a preprint article by Breitwieser and colleagues (n.d.).
These features have convinced several labs to run their simulations using BioDynaMo. Researchers from the University of Cyprus simulate cancer development; scientists from the University of Tel Aviv together with industry partners are working on accelerating drug development; scientists from Newcastle University are studying neural development; and a joint team from the TU Darmstadt and GSI simulate the damage induced by exposure to ionising radiation on the tissue level.
During the recent COVID-19 pandemic, BioDynaMo has been modified to run simulations on how the virus SARS-CoV-2 spreads through a population. Due to its modular design, it was fairly simple to change the agents from having cell to having human behaviours allowing to model different epidemic scenarios, where humans are either Susceptible, Exposed, Infected, or Resistant (SEIR model) (Figure 2). Using an agent-based system allows for the simulation of global models as well as very fine-grained models where the agents are contained in a city, neighbourhood or street.
The conclusions taken from these studies are useful not only to control the virus in the present but also to know how to deal with similar viruses and future outbreaks. In addition, the current pandemic will provide a trove of experimental data that can be used to tune the models and simulations to be more precise next time.
Use Case: COVID-19 Spread in Closed Environments
In one of these types of simulation, BioDynaMo is used to study the spreading of viruses in indoor spaces, specifically SARS-CoV-2 virus that causes COVID-19, in droplets and aerosols. We are investigating several scenarios, such as public transportation (bus, metro) and buildings (supermarkets, offices). In these simulations BioDynaMo is in charge of simulating the behaviour and characteristics of individuals, while the ROOT (Brun and Rademakers 1997) geometrical modeller is used to define the precise environmental geometry (Figure 3).
Each individual can then independently move around in these environments where infected individuals can possibly contaminate healthy ones through the spreading of droplets and aerosols. By studying different geometries, airflows, distancing, masks and other parameters, we can hopefully determine, which environments are best to avoid virus build-up and prevent people from getting infected. This is a work in progress.
This study is done in close cooperation with the epidemiological department of the University of Geneva to make sure that our simulations reflect correctly the many observed cases of virus outbreaks in closed spaces.
This work is sponsored by a grant from the European Open Science Cloud (EOSC) and will be made available as a programme that can be run on the EOSC infrastructure for other scientists’ benefit.
Conflict of Interest