Group 2 – Real Life Applications of Probability Generating

Functions (PGFs)
1) Probability generating functions in infectious disease modeling:
PGFs help predict several properties about early outbreak behavior of infectious diseases in large
populations. These properties include probability of an epidemic, the size distribution as
generations pass and the cumulative size distribution of outbreaks that are non-epidemic.
An important tool for understanding the stochastic behavior of an outbreak soon after
introduction is the probability generation function.
PGFs give insight about the statistical behavior of outbreaks before they are large enough to be
affected by the finite size of the population. We can also analyze and investigate the early growth
rate of disease, the probability the disease becomes established or the distribution of final sizes of
outbreaks that fail to become established.
First application under this topic:
-investigating properties of epidemic emergence in a discrete-time, generation-based framework,
focusing on the probability of extinction and the sizes of outbreaks assuming that the disease is
invading a sufficiently large population with enough mixing so that we can treat the infections
independently.
Second application (discrete time spread) under this topic:
-early extinction probability
In any finite population, a disease will eventually go extinct because the disease interferes with
its own spread. Here, we focus on the question, “What is the probability the outbreak goes
extinct before causing an epidemic?”
Using the PGF we can also find other probabilities like the probability of the population in
question, going extinct. For these we utilize branching processes.
Branching Processes - used frequently to model all sorts of real-world phenomena (population
growth, disease spread, etc.). They model the ‘branching out’ of some sort of
population (for our purposes, let’s think of cells). Each cell lives for one time point
and, at the end of the time point, either reproduces by creating some number of
identical offspring (identical to the original cell) or doesn’t reproduce at all (and
simply dies). This process continues independently with each of the remaining cells; if
all the cells die out without reproducing, nothing else happens (the cells have gone
extinct).
We can formalize this with some notation. Let Xt count the number of living cells in
generation t. We define X0=1, which is essentially saying that there is one cell to start.
The cell then has some ‘generation distribution’ which we will refer to as
the offspring distribution
Every cell that is created after the original cell has this identical generation distribution; also, the
reproductions of every cell is independent of other cells. The population goes extinct as
soon as, well, the population is of size zero. The extinction probability of the process is
the probability that the process ever goes extinct and this can be calculated using the PGF
of a random sum of random variables (the branching process denoted Xt).
https://bookdown.org/probability/bookdown-demo/branching-processes.html

2) The application of probability-generating functions to linear-

quadratic radiation survival curves (Technical Report)
Purpose: To illustrate how probability-generating functions (PGFs) can be employed to derive a
simple probabilistic model for clonogenic survival after exposure to ionizing irradiation.
Clonogenic cell/ Clonogen- a cell that has the potential to proliferate and give rise to a colony of
cells/ cell that has the ability to form clones.
Ionizing- convert (an atom, molecule, or substance) into an ion or ions, typically by removing
one or more electrons.
Irradiation- the process by which an object is exposed to radiation.
Ionizing Irradiation - radiation consisting of particles, X-rays, or gamma rays with sufficient
energy to cause ionization in the medium through which it passes.
Methods: Both repairable and irreparable radiation damage to DNA were assumed to occur by
independent (Poisson) processes, at intensities proportional to the irradiation dose. Also,
repairable damage was assumed to be either repaired or further (lethally) injured according to a
third (Bernoulli) process, with the probability of lethal conversion being directly proportional to
dose.
Using the algebra of PGFs, these three processes were combined to yield a composite PGF that
described the distribution of lethal DNA lesions in irradiated cells.
The composite PGF characterized a Poisson distribution. This distribution yielded the
conventional linear–quadratic survival equation.

The linear–quadratic equation is commonly used to describe the proportion of cells that continue
to replicate after they have been exposed to a dose ionizing radiation. Expressed in mathematical
terms, the fraction of surviving clonogens, S, is related to dose, D, such that

The term in the exponent containing the constant α is often interpreted to depict irreparable
damage; the constant β depicts repairable damage.
Several stochastic models for radiation survival have been derived. These generally employ the
theory of Markov processes. Markovian models, however, can be complicated and difficult to
apply. Simpler stochastic models can be derived using a mathematical tool called a probability-
generating function (PGF). Here, PGFs will used to derive a linear–quadratic survival model.

3) Rumor Spreading of an Susceptible-Infected-Hibernator-Removed

(SIHR) Model in Heterogeneous Social Networks Based on Probability
Generating Function
Purpose: Focuses on the modeling of a rumor spreading in heterogeneous networks. Using the
probability generating function method and pair approximation method, the current research
obtains nonlinear differential equations to describe the dynamics of rumor spreading.
SIHR Model - extends the classical Susceptible-Infected-Removed (SIR) rumor spreading
model by adding a direct link from ignorants to stiflers and a new kind of people – Hibernators.
Heterogenous Network- contains interconnected nodes and links of different types. Such
interconnected structures contain rich information, which can be used to mutually enhance nodes
and links, and propagate knowledge from one type to another
Node- a point in a network or diagram at which lines or pathways intersect or branch.
Propagate - spread/promote/transmit widely or in a particular direction along a medium
Pair Approximation- a method for deriving ordinary differential equations that describe
stochastic, lattice-based dynamics by considering both the frequency of “individuals” as well as
the frequency of pairs of “individuals”. It follows naturally birth processes (like the spread of
plants and animals). It is a way to capture “cellular” interactions (those between individuals in a
discrete sense) using differential equations.
The comparison between numerical simulations and *Monte Carlo simulations confirms the
accuracy of our model. The numerical simulation results show that the heterogeneity of the
network accelerates the outbreak of rumors but reduces the maximum density of spreader and the
scale of rumors.
*Monte Carlo Simulations- aka Monte Carlo Method or a multiple probability simulation, is
a mathematical technique, which is used to estimate the possible outcomes of an uncertain event.
They model the probability of different outcomes in a process that cannot easily be predicted due
to the intervention of random variables. It is a technique used to understand the impact of risk
and uncertainty in prediction and forecasting models.
Heterogeneity - the quality or state of being diverse in character or content.