Dynamic Modeling and Simulation for Molecular Communication Networks

Date: 17th – 18th September, 2018.
Location: 6 Cromwell Place, London SW7 2JN.
Organizers: Malcolm Egan (Inria and INSA Lyon) and Trung Q. Duong (QUB).
Participants:
Summary
On the 17th and 18th of September 2018, the seed workshop “Dynamic Modeling and Simulation for Molecular Communication Networks” was held at the Embassy of France in London. The workshop involved researchers from Inria, CNRS, Queen’s University Belfast, Warwick, King’s College London, Liverpool, Waterford, Perugia, the Spanish Research Council and Graz.
The meeting focused on molecular communication, both in terms of applications and theory. A highlight of the meeting was the discussion on interactions between information theory, signaling in biological systems, and reaction-diffusion models. It also revealed a healthy multidisciplinary community in Europe interested in molecular communication from an information processing perspective.
Programme
Day 1
12h00: Arrival and lunch.
14h00 – 14h30: Introduction
14h30 – 15h00: Yansha Deng (KCL)
15h05 – 15h35: Irene Otero-Muras (SRC)
15h35 – 16h10: Discussion Break
16h10 – 17h00: Mauro Femminella (Perugia)
17h00 – 17h45: Discussion Break
17h45 – 17h15: Daniel McGuinness (Liverpool)
19h00: Dinner
Day 2
09h00 – 09h30: Michael Barros (Waterford)
09h35 – 10h05: Adam Noel (Warwick)
10h10 – 10h40: Discussion Break
10h40 – 11h10: Bao Quoc Tang (Graz)
11h15 – 11h45: Malcolm Egan (Inria)
11h45 – 12h15: Discussion Break
12h15 – 13h00: Open Discussion
13h00 – 14h30: Lunch
14h30 – 15h30: Open Discussion
15h30 – 16h00: Summary and Close of Workshop
Presentations

Presenter: Weisi Guo (Warwick)
Title: Molecular Communications with Vortex Rings
Abstract: Vortex rings are self-propagating structures that can retain coherent shape throughout its propagation life time, dramatically reducing ISI over long distances and enabling a high symbol rate. These highly directional structures can propagate in turbulent conditions for up to 50 to 100 times the nozzle diameter without guidance or drift. We show both theoretical underpinning equations, simulations, and experimental results. However, several research challenges exist in generating reliable vortex ring sequences for communication purposes and we need both classical communication techniques and new cross-disciplinary methods to overcome them.

Presenter: Yansha Deng (KCL)
Title: Internet of Bio-Nano Things [Slides]
Abstract: This talk will introduce the emerging field of molecular communication wherein chemical signals are used to connect “tiny” machines such as living cells, synthetic biological devices and swarms of microscale robots. It begins with an overview of molecular communication systems and how they are modeled; each has a Transmitter, the Propagation Channel, and the Receiver, just as in a conventional communication system. She will describe the modelling and simulation of a diffusive molecular communication system with a reversible adsorption receiver in a fluid environment. She will explain how to analytically model the time-varying spatial distribution of information molecules and present a simulation framework for the proposed model that accounts for the diffusion and reversible reaction. She will also cover her talk on signal processing via chemical circuits-based microfluidic design.

Presenter: Irene Otero-Muras (SRC)
Title: Biochemical Reaction Network Analysis for Cell Signaling and Gene Regulation [Slides]
Abstract: Signal transduction and gene regulation are crucial in how cells perceive and respond to their environment. Modeling signaling cascades and gene regulatory networks within the formalism of biochemical reaction networks helps understanding these mechanisms from a system’s perspective, which is important for applications in systems and synthetic biology.

Biochemical reaction networks exhibit a rich variety of complex nonlinear dynamic behaviours, including bistability, limit cycle oscillations or chaotic behaviour. We are interested in understanding the origin of this complex behaviour, in particular bistability, both from a thermodynamic perspective (entropy balance) and from a purely structural (CRNT) approach. Bistability is important in biology, as it confers cells with the capacity to “make decisions”, as it occurs in apoptosis, stem cell differentiation or differential cytokine signaling.

As an example of application in systems biology, our methods for bistability detection allowed us to elucidate the molecular mechanisms responsible for the differential signaling in type 1 interferons (explaining how coding through the same ligand, receptor and signaling pathway, interferon leads to apoptotic or antiviral activities depending on the interferon dose and/or the stability of the ligand-receptor complex at the cell membrane).

Regarding the engineering of biocircuits, combining biochemical reaction network modeling with mixed-Integer Nonlinear optimization we have developed a framework for the automated design of biochemical reaction networks with target pre-defined behaviours, that we used for example to design synthetic oscillators and switches, and also to find gene regulatory motifs with pre-defined functionalities (including biochemical adaptation, change fold detection, or pattern formation). Since gene regulatory networks are inherently stochastic, in future works we aim to take into account the effect of molecular noise in the design of biocircuits.

Presenter: Mauro Femminella (Perugia)
Title: Research Activity in Molecular Communications at University of Perugia [Slides]
Abstract: The team of University of Perugia (currently consisting of M.
Femminella, L. Felicetti, and prof. G. Reali) has been actively
contributing to the research on molecular communications since 2011. The
main activity has been the development of a simulator, named BiNS and
currently at version 2, which is able to simulate the movement and
interaction between (biological) nanomachines and released molecules.
This simulator, implemented in Java, is characterized by high
scalability features, obtained by means of multi-thread programming and,
recently, GPU offloading of the most computing intensive task.
Additional activity carried out by the team have been system modeling,
especially modeling of receptor-based receivers by means of queueing
theory, and communication protocols design, carried out by defining a
finite state machine for each node involved in the communication. The
talk ends with a discussion about past experience in projects funded by
EU (project CIRCLE) and locally by University of Perugia, and future
outlook of funding.

Presenter: Michael Barros (Waterford)
Title: Bio-Nano-Machines Communication Systems Inside Cellular Tissues and Their Role in Precision Medicine
Abstract: Since the early days of life on earth, bio-nano-machines have been organising themselves with communications mechanisms that allowed the evolution from unicellular systems to multicellular systems. It is now well known that multiple multicellular systems are co-dependently organised to become entire living organisms, and they constitute the main pillar for human life to exist as we know. However, it is not only till recently that we started to understand how this type of communications in the same way we understand communications engineering, named Molecular Communications. This field is composed of a plurality of systems that use molecules as information carriers such as the communication inside cellular tissues. Now, bio-nano-machines communication can be not only further understood but engineered to performed increased sensor and actuation tasks. At the same time, the many diseases that alarm world society nowadays are often linked to failures in those communicating biological machines. With this vision, we can use the existing various communication engineering theories and methods to control these communication systems towards developing novel diagnosis and treatment of diseases in the cellular scale, which is the main vision of Precision Medicine. In this talk, I will introduce how these multicellular mechanisms are organised through communications, how diseases can start as molecular communications failures (e.g. Alzheimer’s, Parkinson’s, and so on), how communication engineers are increasingly contributing to this new research topic, and how molecular communications can create a tremendous impact on Precision Medicine. It is expected that both the advancements in nanotechnology and in molecular communications can one day be finally merged towards the Internet of Bio-Nano-Things, and revolutionise the society we living in.

Presenter: Adam Noel (Warwick)
Title: Efficient and User-Friendly Simulation of Reaction-Diffusion Systems with AcCoRD [Slides]
Abstract: Molecules are ubiquitiously used in the human body and other biological systems for signaling, control, and regulation. From the perspective of a communications engineer, these molecules carry information. Molecular communication (MC) is an emerging multi-disciplinary field that applies communications engineering tools to molecular signalling. This field seeks to improve our understanding of biological systems and build networks of devices that could operate in fluids. This talk will present AcCoRD (Actor-based Communication via Reaction-Diffusion), which is an open source “sandbox” simulation tool to make the simulation of molecular reaction-diffusion dynamics more accessible to the communications engineering research community. The talk will demonstrate use of the simulator while focusing on the development of recent features, in particular the efficient simulation of absorbing surfaces and an algorithm to add flow to simulations in the mesoscopic regime.

Presenter: Bao Quoc Tang (Graz)
Title: Large time behaviour of reaction-diffusion systems modelling chemical reaction networks [Slides]
Abstract: This talk presents the quantitative large time behaviour of chemical reaction networks which are modelled by reaction-diffusion systems. We analyse first-order reaction networks which result in linear reaction-diffusion systems. By utilising the entropy method, we show exponential convergence to equilibrium for weakly reversible networks. When a network is not weakly reversible, by using topological ordering, it can be partitioned into {\it source}, {\it transmission} and {\it target} components, whose behaviours can be determined completely by an extension of the entropy method. When some substances of a network do not diffuse, we investigate the {\it indirect diffusion effect}, i.e. the combination of diffusive substances and (weakly) reversible reactions lead to some “diffusion effect” of non-diffusive substances. Extensions to high-order chemical reaction networks, i.e. nonlinear reaction-diffusion systems, are also discussed.

Presenter: Malcolm Egan (Inria and INSA Lyon)
Title: Coexistence in Molecular Communications [Slides]
Abstract: Some of the most ambitious applications of molecular communications are expected to lie in nanomedicine and advanced manufacturing. In these domains, the molecular communication system is surrounded by a range of biochemical processes, some of which may be sensitive to chemical species used for communication. A key question is therefore what impact the communication system has on these processes. In this talk, we formalize this question within an information theoretic framework and establish a formal connection to the problem of covert communication. This provides a means of obtaining fundamental information theoretic limits of molecular communication in or nearby biological systems.

Presenter: Daniel McGuinnes (Liverpool)
Title: Macro-Scale Molecular Communications [Slides]
Abstract: Molecular communication (MC) is a method where the information transmission involves the use of chemicals or molecules instead of electromagnetic (EM) waves. The study of MC to date has mostly focused on the nano to micro scale and there have been only a few experimental studies on the subject of macro-scale. In this presentation, the properties of macro-scale MC are experimented and studied. These are signal flow (q), carrier flow (Q) and bit duration (T). A mass spectrometer (MS) with a quadrupole mass analyzer is used as the detector and an in-house-built odor generator is used as the chemical pulse generator. The study has shown that the signal energy has a quadratic relation to signal flow whereas the carrier flow has a non-linear relation to both the signal amplitude and the signal energy. The bit duration has shown that the system reaches a saturation point as the bit duration is increased and this effect also occurs in the leftover chemical signal after bit transmission. Finally a mathematical model is developed for the first time to explain the molecular transmission on the macro-scale using advection-diffusion equation.