This model evaluates the effect of random SARS-CoV-2 variants on morbidity of COVID-19 in a susceptible population. New virus variants are constantly being produced through genetic mutations, which can result in differences in virulence. It is hypothesized that if a random mutation increases the number of viral agents, then a higher percentage of the susceptible population will become infected, thus increasing the morbidity of SARS-CoV-2.

This model used two agent types, one type representing humans and the other type representing the SARS-CoV-2 virus. Infected human agents used 5 times more step energy than uninfected agents to simulate the effects of COVID-19. The experimental scenario used random mutations in the virus agent that were generated by mutating breed energy. This led to a change in population of COVID-19 agents, changing how likely it was for human agents to get infected. The control scenario did not use random mutations in the virus agent. The total number of ticks that it took for human agents to become extinct as a result of being infected in each scenario was compared.

Half the trials indicated that the mutated COVID-19 variant was more virulent than the control, while the other half indicated that the variant was less virulent. Therefore, there is no relationship between mutations in the virus agent and morbidity in the human population. Random SARS-CoV-2 variants have the potential to be equally either more or less virulent.

​

Model Instructions

Shifting cultivation (also referred to as swiddening or slash-and-burn) is an agricultural system commonly practiced in tropical and subtropical regions that can be spatially represented in a simulation model. This research examined the boundary dynamics of deforestation in an area under shifting cultivation using Complex and Organized Behaviour Within Environmental Bounds (COBWEB), an agent-based modelling software. A model was developed to demonstrate how rates of deforestation would shift with particular system changes, namely crop yields and human population growth. Simulations revealed that humans are inextricably dependent upon trees, as there will always be a non-zero rate of deforestation present, but the ability to extract energy from crops does not necessarily remove reliance upon energy derived from trees if human population growth is high. As such, the rate of human population growth is a better indicator of deforestation than crop yields, and under the particular parameters and design of this model, shifting cultivation is found to be sustainable. While deforestation dynamics in shifting cultivation systems have been robustly modelled in related scholarship, this research provides a simplified, agent-based approach to modelling that can be adapted to further explore system changes.

Modeling and simulating the relationship between intergroup conflict and intragroup coordination on COBWEB simulation program.

According to the statistics from the Government of Canada, roughly 2.4 million people were diagnosed with asthma by a health professional in 2014. Aside from genetic predisposition, environmental allergens such as PM are one of the most common risk factors of asthma. However, we haven’t brought our attentions to the underlying mechanism of asthma. Understanding the biological process of a disease is crucial in developing effective medical treatments. Recent studies have found that dysregulation in autophagy is associates associated with critical clinical features of asthma such as airway inflammation, mucus hyperproduction and airway remodeling.

Autophagy has not only been related to disease such as cancer, neurological disorders, and respiratory diseases, but it has also gained a fair amount of attentions in nutritional science. A popular eating pattern called intermittent fasting would activate regulated autophagy which in turn removes the wastes in our body for weight loss and health-related purposes. Thus, a simulation of the autophagy process in COBWEB was modulated to enlighten this biological pathway to people either with some scientific backgrounds or not.

This project was an attempt at offering a visual representation of hypothalamic-pituitary-adrenal (HPA) axis dysfunction, in the form of hypercortisolism, in neurotypical individuals and individuals with Autism Spectrum Disorder (ASD) respectively. Additionally, it aimed to discuss the potential relationship between HPA axis dysfunction and the high rates of depression observed in the ASD community.

​

This was a summer project and involved the following high school students: Priscilla Fung, Stephie Liu, Janelle Liu, Quynh Hai MyNguyen, Olakunsi Peters, Sudipta Saha, Druphadi Sen, Sneha Shah and Aishwarya Sudarshan.

Recent advances in computer technology have brought a revolution in ecological modelling. Ecoinformatics and computational ecology make use of various programs, including agent-based modeling algorithms, to study ecological systems. In this study, an in-silico analysis was performed using an agent based modelling software, to analyze the impacts of a potential toxin on Lake Ontario ecosystem. For easier duplication of the real world into the virtual system, the ecosystem was divided into 6 compartments. These compartments include phytoplankton, zooplankton, macroinvertebrates, forage fish, piscivores, and sea lamprey. The test model was performed under five different concentrations of toxin. Each test was repeated 15 times to reduce demographic stochasticity. The results suggest that toxic contaminations, such as mercury, could potentially lead to population reduction in forage fish, piscivores and sea lamprey compartments.

Les progres recents relies a la technologie informatique ont amene une revolution dans la modelisation ecologique. L’eco-informatique et l’ecologie computationnelle utilisent plusieurs programmes, y compris des algorithmes bases sur les systemes multiagents pour etudier les systemes ecologiques. Dans cette etude, une analyse in silico a ete accomplie en utilisant les systemes multiagents pour analyser les impacts d’une toxine potentielle dans l’ecosysteme du Lac Ontario. Afin de mieux ameliorer la representation du monde reel dans le systeme virtuel, l’ecosysteme du Lac d’Ontario a ete divise en six compartiments. Ces compartiments comprennent le phytoplancton, le zooplancton, les macroinvertebres, les poissons fourragers, les piscivores et la lamproie marine. Ce modele a ete examine sous cinq concentrations des toxines differentes. Chaque examen a ete repete 15 fois pour reduire la stochasticite demographique. Les resultats suggerent que des contaminations toxiques, comme la contamination par le mercure, pourraient potentiellement arriver a une reduction de la population des poissons fourragers, des piscivores et des compartiments de la lamproie marine.

This COBWEB project simulates the PI3-K/Akt insulin signalling pathway but may serve as an base model for all signal transduction pathways. This project visually demonstrates signal amplification using the disease and waste tabs: the COBWEB disease function depicts the phosphorylation cascade from agent to agent, and the waste function allows for uptake of substrates in the extracellular domain (abiotic zone 1). The behaviours of each agent are not specific; thus, it can be adapted to model any signalling pathway by changing the proportions of each substrate (amount of each agent, number of agents) and the order of cell communication steps (how one agent transfers disease to the next agent).

This model explores the effect of higher density living areas on the transmission of COVID-19 in a university residence. Higher density living areas can be defined as locations where people spend time with others. It was hypothesized that the introduction of higher density regions in a constant overall population density will lead to quicker spread of COVID-19 because it will increase how often individuals come in contact with each other.

A university residence was drawn with stones on the COBWEB grid and 100 agents were used to represent students living in the residence. Two scenarios were compared: one with higher density living areas with twenty-five dorm rooms and a common area, and the control scenario using the total area of the residence. Each trial started with one infected individual. For the higher density living areas scenario, all agents were placed in the common area during the day and divided among the dorm rooms at night. The process was repeated until all agents were infected. The total number of ticks that it took for all agents to be infected in each scenario was compared.

The results validate the hypothesis because it took 208 less ticks for all agents to be infected with the introduction of higher density regions compared to the control, showing that there is a relationship between higher density living areas and increased COVID-19 transmission. Individuals in higher density regions are more likely to get infected because of their close contact with others.

Using COBWEB, I modelled the effects of particulate matter within alveolar epithelial tissue. In short, I was looking at how reactive oxygen species are created in lung tissue from breathing in particulate matter (using Fenton type reactions) and how these reactive oxygen species can affect tissue. To do this, I created two pneumocytes (type I and II), sent in particulate matter to the system, monitored how the particulate matter reacted to form reactive oxygen species, and then also monitored how these reactive oxygen species destroyed tissue.

Background: Psoriasis is a chronic inflammatory skin disease characterized by red and scaly patches on the epidermis. It results from overactive immune responses induced by the NF-κB protein complex, which increases proinflammatory gene expression. As a result of increased inflammatory responses, keratinocytes in the epidermis are overproduced and accumulate, creating psoriatic plaques. One cause of this overactive immune response is attributed to genetic mutations in the caspase recruitment domain-contain protein 14 (CARD14), which causes overactivity of NF-κB via amplified enzymatic activity of the mucosa-associated lymphoid tissue lymphoma translocation protein 1 (MALT1). Topical therapies for psoriasis do not significantly alleviate symptoms for severe cases, however, the systemic inhibition of MALT1 may be more effective. One potential therapeutic compound is mepazine acetate, a phenothiazine derivative that allosterically inhibits MALT1.

Methods: The impact of CARD14 mutations on the development of psoriasis was modelled using the COBWEB simulation software. COBWEB was then used to model the inhibition of MALT1 by mepazine acetate by comparing the control with treatment models.

Results: The simulations with the mepazine acetate treatment had a reduced accumulation of dead keratinocytes and decreased proliferation of psoriatic plaques in the upper epidermal layers, compared with control simulations after seven days.

Limitations: This model is an idealistic representation of psoriasis; variation between these results and clinical findings is likely.

Conclusion: These results demonstrate that mepazine acetate decreases MALT1 activity, thus reducing proinflammatory responses. The next steps include building a more complex model with the ability to substantially vary responses to treatment.

Our group investigated the effectiveness of handwashing, masks, and gloves in minimizing community transmission of COVID-19 when 30% and 50% of the population abides with each of these precautions. COBWEB 2D was used to model the transmission rate of the virus in the city of Toronto, Canada.

​

Additional students involved in this project include: Pooja Dey, Christina Devarapalli, Selim Choi, Heather Li and Sirat Munjal.​