simulationeverywhere-models / cadmium-seir-asyn

SEIR model with asymptomatic and quarantine

SEIR-Asyn

The model is based on the following paper:

Tang, B., Wang, X., Li, Q., Bragazzi, N. L., Tang, S., Xiao, Y., & Wu, J. (2020).
Estimation of the Transmission Risk of the 2019-nCoV and Its Implication for
Public Health Interventions. Journal of clinical medicine, 9(2), 462. https://doi.org/10.3390/jcm9020462

This instructions assume that:

1. Cadmium has been installed following the instructions in the manual: http://www.sce.carleton.ca/courses/sysc-5104/lib/exe/fetch.php?media=cadmiumusermanual.pdf

2. The SEIR-Asyn model is cloned inside the folder: Cadmium-Simulation-Environment/DEVS-Models (see the installation manual)

Compile the model

1. Install Cadmium following the instructions in the manual: http://www.sce.carleton.ca/courses/sysc-5104/lib/exe/fetch.php?media=cadmiumusermanual.pdf

2. Clone the repository inside the folder: Cadmium-Simulation-Environment/DEVS-Models https://github.com/SimulationEverywhere-Models/Cadmium-SEIR-Asyn

3. Compile the model using the make file

Model implimentation

Parameters used in the model

(These parameters can be modified following the instructions below)

Symbol	Symbol in paper	Value from paper	Description
c	c	14.781	Contact rate per person per day
b	β	2.1011e-8	Probability of transmission per contact
q	q	1.8887e-7	Probability of being quarantined per contact
e	σ	0.142857	Proportion of exposed individuals becoming infective per day
l	λ	0.071428	Proportion of quarantined uninfected individuals being released per day
n	ϱ	0.86834	Probability of having symptoms among infected individuals
di	δI	0.13266	Proportion of symptomatic infective individuals going into quarantine per day
dq	δq	0.1259	Proportion of quarantined exposed individuals progressing to quarantined infective per day
yi	γI	0.33029	Proportion of symptomatic infective individuals recovering per day
ya	γA	0.13978	Proportion of asymptomatic infective individuals recovering per day
yh	γH	0.11624	Proportion of quarantined infective individuals recovering per day
a	α	1.7826e-5	Proportion of symptomatic infective and quarantined infective individuals dieing per day
t	θ	0	How infective are asymptomatic infective people compaired to symptomatic infective people
S	S	11081000	Initial susceptible population
E	E	105.1	Initial exposed population
I	I	27.679	Initial symptomatic infective population
A	A	53.839	Initial asymptomatic infective population
Sq	Sq	739	Initial quarantined susceptible population
Eq	Eq	1.1642	Initial quarantined exposed population
H	H	1	Initial quarantined infective population
R	R	2	Initial recovered population
D	D	0	Initial deceased population

Differential equations used in the model

Class	Equation in paper	Equation in implementation
S'	-(b×c+c×q×(1-b))×S×(I+t×A)+l×Sq	-(c×(I+A)×S×b×(1-q)) - (c×(I+t×A)×S×b×q) - (c×(I+t×A)×S×(1-b)×q) + (l×Sq)
E'	b×c×(1-q)×S×(I+t×A)-e×E	-(e×n×E) - (e×(1-n)×E) + (c×(I+t×A)×S×b×(1-q))
I'	e×n×E-(di+a+yi)×I	-(di×I) - (yi×I) - (a×I) + (e×n×E)
A'	e×(1-n)×E-ya×A	-(ya×A) + (e×(1-n)×E)
Sq'	(1-b)×c×q×S×(I+A)-dq×Eq	-(l×Sq) + (c×(I+A)×S×(1-b)×q)
Eq'	b×c×q×S×(I+A)-dq×E	-(dq×Eq) + (c×(I+A)×S×b×q)
H'	di×I+dq×Eq-(a+yh)×H	-(yh×H) - (a×H) + (di×I) + (dq×Eq)
R'	yi×I+ya×A+yh×H	+(yi×I) + (ya×A) + (yh×H)
D'	[not directly modeled]	+(a×I) + (a×H)

Run the model

OPTION 1 - Default parameters

./bin/SEIR_Asyn (linux)

./bin/SEIR_Asyn.exe (windows)

OPTION 2 - Define your own parameters

Place a text file (e.g. my_data.txt) with the input parameters in the folder input_data

Run the program normally using

./bin/SEIR_Asyn my_data.txt (linux)

./bin/SEIR_Asyn.exe my_data.txt (windows)

NOTE Follow the instructions in input.txt for how to lay out your data

Visualize the results

runner.py and plotter.py are python3 files that assume that you are in a linux like environment.

runner.py contains the functions required to run the model repeatedly and in parallel with programatic arguments, and is meant to imported

plotter.py contains the functions needed to read in the output of the model, and an example of how to use runner.py and mathplotlib to generate line graphs from the output