Analyzing heterogeneous spreading dynamics

People involved:
Liubov Tupikina
Marc Santolini
Aurelie de Faure
Gael Simon

Keywords: spreading, diffusion, heterogeneity, attributes, networks, epidemics, information

Project description

The theoretical part of the project is analysis of spreading dynamics on heterogeneous networks. Why heterogeneous? Because most of the real world networks are intrinsically heterogeneous: transportation networks with different width, biological fractal network of human lungs etc. The main question is how to characterize the influence of the heterogeneity on the transportation phenomena. This includes designing the network indicators for characterizing complex dynamics, developing the fundamental framework applicable for disease spreading processes. In this project we are working on designing and testing heterogeneous epidemics spreading models inspired by the real data.
The practical aspect of this project is then to dissect the influence of the heterogeneity from the other factors and to incorporate physical, epidemiological and socio-economic mechanisms into the heterogeneous spreading models. The most existing part of the project is actually implementation of the indicators to such complex models in order to detect the localisation of epidemics spreading and to identify influential spreaders even in the heterogeneous setup. For more information about the networks please visit  here.


Innovative educational methods for research integrity

People involved:
Aida Bafeta
Ariel Lindner

Keywords: ethics, research, misconduct, serious games

Project description

Research misconduct and misbehaviors are perceived as a risk of weakening the institutions and the scientific community. Why would the public fund research and trust its researchers if they are not honest?  In the media sphere, fraud, data fabrication and falsification, and plagiarism receive a particular public attention, but according to a meta-analysis of survey data only 2% of researchers admit to having used these practices at least once. While  a third of scientists admitted to having used questionable research practices, such as modifying the design, methodology, or results of a study in response to pressure from a funding source. Today, it is important to rethink the concept of integrity research beyond of misconduct (fraud, data fabrication and falsification,  and plagiarism). One of the proposed solutions is to educate scientists on research integrity. Improved research integrity training is now variously supported and mandatored. However most of studies evaluating research integrity training have mostly been inconclusive, long-term impacts on behavior have not been demonstrated, and focused on the  plagiarism. It is crucial to identify questionable research practices and to propose training for researchers. We believe that developing simple, playful and adequate educational resources on questionable research practices in the form of serious games, could help to solve this problem. In this project we will want to create several tools dedicated to early researchers and students to raise awareness of the integrity of research.

Physical, blockchain-mediated multiplayer games of trust

People involved: Asa Calow 

Keywordshardware, blockchain, multiplayer, system dynamics

Project Description

Blokc is a proposed low-cost and general purpose hardware/software ecosystem for blockchain research in physical multiplayer environments –allowing for the testing of multi-agent system models with co-located participants, in dynamic and trust-free environments.

Blokc will go beyond the existing state-of-the-art for embedded blockchain technology – e.g. open source Bitcoin hardware wallet Trezor – by building out the underlying core technologies and libraries for common hardware platforms such as Arduino and Particle Photon, as well as laying out fundamental principles of design when combining pre-determined rules for interaction as described by smart contracts (deployed via e.g. the public Ethereum blockchain, or a private instance of JP Morgan Chase’s Quorum platform) with custom hardware such as barcode readers, internal geolocation, gesture recognition, and Bluetooth Low Energy or Mesh. These design principles will be backed up by a well-documented suite of open source tools, allowing other researchers to extend and build new platform variations to suit their own needs.

Specific technical challenges are expected to include the investigation of key blockchain protocols in an embedded context: Light Ethereum; scalable autonomous smart contracts (Plasma); proof-of-stake; and non-interactive zero-knowledge proofs (Zerocoin’s zk-snarks).

Beyond these technical underpinnings, the Blokc project will set out to conduct live multiplayer experiments with new blockchain-mediated “trusted interaction” models – to foster co-operation and resource sharing, or build shared provenance and value networks for instance – with immediate application in today’s dynamic and trust-free environments such as disaster zones, refugee camps, or temporary cities (e.g. Burning Man).

The final part of the research involves blockchain data replay, analysis, and visualisation. This would enable would-be dynamic systems designers to gain insight on whether/how game-theoretic models of human collaboration differ once deployed in physical environments with live actors.

Blokc is part of a larger piece of art-science-technology work – the Institute of Unknown Purpose, a collaboration with Professor James Crutchfield, director of UC Davis’ Center for Complexity Science. The IoUP is engaged in simultaneously speculating about and designing foundational technologies for a post-Transition world. Beyond this most explicit link to the call theme, the project links directly to several of the topics discussed during the kickoff workshop – new tools for exploring differences between proposed game theoretic and live experimental environments; new forms of digital governance; methods for designing accountability.

Dissecting science

People involved: Marc Santolini

KeywordsScience of science, collective intelligence, team network, Open science, science of success

Project Description

With the advent of digital tools to produce and share knowledge, Science is undergoing a deep transition towards a more decentralized and inclusive framework. This transition will be shaped by the integration of machine learning and artificial intelligence, allowing to build efficient recommendation systems and find needles in a stack of massive amounts of open knowledge through designed serendipity. For such a design to operate as efficiently, transparently and ethically as possible, scientists need to turn the microscope on themselves and understand the scientific process itself. This endeavour, which has for long been the realm of the philosophy of science, has recently gathered a new crowd of physicists, mathematicians, statisticians or social scientists: this is the advent of the Science of Science. Researchers are linked by co-authorships, ideas by citations, research institutions by collaborations, forming networks at various scales that underlie the scientific process. Such networks are not only a convenient description, they are also predictive of the success or percolation of an idea, a paper, a collaboration. In this project, we aim to dissect Science further and understand its underpinnings in situ by using the iGEM scientific competition as a “model organism”. For more than 10 years, 2000+ participating student teams (from high school to graduate) have worked during the summer on synthetic biology projects and have been awarded medals and prizes. Each team documents their work in an open Lab Notebook, allowing to capture features that associate with their end success. While this offers insights on the process of doing science, it remains a virtual footprint of complex real-world interactions. Here, we will overcome this issue by following the face-to-face interactions in the lab for a significant number of teams through discreet sensors. This will offer unprecedented measurements of  the dynamical processes underlying the making of science, exhibiting the anatomy of Science in the making to better understand and design its transition

For more info visit

Genomic signatures to identify non-tubercular Mycobacterial species

People involvedAnshu Bhardwaj 

Keywordsgenomics, biomarker, pathogen, nontuberculous mycobacteria, AI chatbot, gaming

Project Description

Antimicrobial resistance (AMR), the phenomenon of clinically relevant pathogens developing multi-drug resistance (particularly to antibiotics), has emerged as a grave threat to public health that could plunge the world into a ‘post-antibiotic era’. Neglecting AMR would result in global annual loss of 10 million lives and trillions of dollars by 2050 (2015 O’Neil). The Global Action Plan for AMR has identified five objectives to address this scourge. The first two objectives – “ (a) Improve awareness and understanding of antimicrobial resistance through effective communication, education and training and (b) Strengthen the knowledge and evidence base through surveillance and research” are the basis of the current proposal which aims to address these using genomics and artificial intelligence tools. At the core of the project is the pressing need to identify infections from Nontuberculous mycobacteria (NTM) or Mycobacterial Other Than TB (MOTT). NTMs include more than 160 ubiquitous Mycobacterium species that do not cause tuberculosis or leprosy. NTMs are present in the environment (water or soil) and can infect humans or animals leading to a range of pathological conditions like pulmonary, skin, bone, joint, and disseminated diseases. NTM species are gaining visibility due an increasing number of strains responsible for treatment-resistant diseases. NTMs are taxonomical diverse and increasing number of new species offer challenges in identification of NTMs in clinical settings. NTM species like Mycobacterium abscessus are now recognized as a major threat and FDA identified NTMs as their focus disease area in 2016-17. Despite their increasing role in human diseases (in India prevalence rate increased from 0.7% to 34%) limited data is available to delineate and identify species of NTMs in the clinical settings. It is crucial to know the NTM species for prescribing treatment options as the disease presentation and clinical investigation parameters are very similar to Tuberculosis. Moreover, there are species-specific differences from context of resistance to different antibiotics. The current proposal entitled ‘G-MOTT’ aims to develop a comparative genomics pipeline to identify genomic signatures that may help in delineating different NTM species. The second part of the proposal aims to build a conversational artificial intelligence (AI) chatbot wrapped as a gaming application. The idea behind building this tool is to engage the mobile users into playing a game to understand the concept and challenges of AMR. This is first of its kind unique effort to utilize the power of AI to strengthen the human-machine interface aligning with the goals of the Global Action Plan on creating awareness, education and training on AMR.