Since the COVID-19 pandemic hit, the Julia language has been used to simulate a wide range of epidemiology applications. These include vaccination impact, hospital utilization, silent transmission impact, effects of early reopening and virtual virus spread through Bluetooth tokens. Here is guide showing different epidemiology studies performed using the Julia language.
This study analyzed the effect of isolation and early symptom identification on the number of ICU beds which would be required at the peak of the COVID-19 outbreak. At the core of this study was an age stratified compartment model simulated in Julia.
This study concluded that vaccination can have a substantial impact on mitigating COVID-19 outbreaks, even with limited protection against infection. At the core of this study was a Julia based agent-based model parameterized with US demographics and age-specific COVID-19 outcomes.
(3) An examination of school reopening strategies during the SARS-CoV-2 pandemic | Paper
This study concluded that surveillance programs using less sensitive tests may be adequate in monitoring infections within a school community by both keeping infections low and allowing for a longer period of instruction. Their model predicted that reducing the number of contacts through cohorts produces a larger effect than diminishing transmission rates per contact. At the core of this study was a Julia written age stratified SEIR (Susceptible-Exposed-Infected-Recovered) model.
This study found that the majority of incidences may be attributable to silent transmission from a combination of the presymptomatic stage and asymptomatic infections. They quantified the effect of isolating silent infections in addition to symptomatic cases, finding that over one-third of silent infections must be isolated to suppress a future outbreak below 1% of the population. At the core of this study was a stochastic, age-stratified agent-based computational model.
This study concluded that, at the peak of COVID-19 outbreaks, the need for ICU beds will exceed the total number of ICU beds even with self-isolation at 40%. They demonstrated the coming challenge for the health care system in Canada and the potential role of self-isolation in reducing demand for hospital-based and ICU care. At the core of this study was a Julia written stochastic, age stratified agent based computational model.
This study was the first Scientific Machine Learning (SciML) model released as the pandemic hit. It augmented the SIR model by integrating neural network components to model the quarantine strength. The resultant model (called the QSIR model) is not only much more expressive than the SIR model, but also highly interpretable in nature! The QSIR model can be used as a diagnostic tool to understand the on-ground quarantine situation in any region of interest.
This study showed that if the Southern and the West-Central US states had not reopened early during June-August 2020, a large number of infections would have been reduced. At the heart of this study is a Scientific Machine Learning (SciML) framework written in Julia.
This study introduced the concept of “SafeBlues” strands. Safe Blues strands are safe virtual “virus-like” tokens that respond to social-distancing directives similarly to the actual virus. However, they are spread using Bluetooth and are measured online. They showed that real-time data on the Safe Blues tokens can be used for estimation of the epidemic’s current and near-future state. The deep learning module in this study is a Julia written Scientific Machine Learning (SciML) framework.
This study showed that transmission-interrupting strategies become relatively more effective with time as herd immunity builds. The study concluded that the most effective vaccination strategy for reducing mortality due to COVID-19 depends on the time course of the pandemic in the population. For later vaccination start dates, use of SARS-CoV-2 vaccines to interrupt transmission might prevent more deaths than prioritising vulnerable age groups. The core of this study was an age stratified SEPAIR (susceptible, exposed, presymptomatic, asymptomatic, symptomatic, removed) model.
(10) Some other tutorials
(a) Various implementations of the classical SIR model in Julia | Code |
(c ) Introduction to Universal Differential Equations: How to integrate neural networks with epidemic models in Julia | Code |
(d) Covid-19 application: Using neural network assisted epidemic models to diagnose quarantine strength in any given region of the world | Code |
(11) Relevant Julia packages
(a) Introduction to Differential Equations in Julia [Documentation]
(b) Introduction to the Scientific Machine Learning (SciML) ecosystem in Julia [Documentation]
(c) Introduction to ModelingToolkit: Language for high performance symbolic computational in Julia [Documentation]
(d) Introduction to Catalyst: A symbolic modeling language for chemical reaction network simulation [Documentation]