Welcome to Become a Disease Detective ! 🔍

Organised by

R Ladies+ Melbourne

Acknowledgement of Country

We would like to begin by acknowledging the Wurundjeri people of the Kulin Nation, the Traditional Custodians of the land on which we gather here today. We pay our respects to their Elders past, present, and emerging, and extend that respect to all Aboriginal and Torres Strait Islander peoples. We recognize their continuing connection to land, waters, and culture, and acknowledge that they were the first scientist on this land.

We are…

Dionne and Kirsty and we are from

R Ladies+ Melbourne.

We are here to teach you how to code in R!

What is coding and what is R ?

Coding is writing instructions that a machine or computer can understand to perform a task.

R is coding language used in statistics, data science and many other disciplines. It is also very powerful tool for visualising data.

Why should you learn R?

⚖️ R is free, open source and everyone can access

🥳 It can do so many things

📚 Lots of resources to learn and perform tasks, with a very large user community

🧑‍🎓 You can get a super cool job ! It’s a fantastic skill that is highly sought after in the work force

What will you learn today?

🗺️ What is an infectious disease, what is an outbreak and who are disease detectives?

🔍 How can R be used to investigate outbreaks ?

Then YOU! are going to solve an outbreak !

What are infectious diseases?

🦠 Infectious diseases are caused by certain microorganisms, such as bacteria, viruses, parasites or fungi, and will make people (or animals) sick when they are exposed to these microorganisms.

💭 Most microorganisms don’t make us sick, but some do, and these are known as pathogens.

🤧 There are lots of different ways that you might get an infectious disease. Some ways include breathing respiratory droplets, eating contaminated food or water, insect bites and many many more !

Pause after reading the headline and see what students have to say

When do infectious diseases cause outbreaks?

📚The World Health Organisation definition of an outbreak:

💡 “occurrence of disease in excess of normal”

This might be because of a new pathogen, bringing pathogens to new areas, a mutating pathogen.

Many infectious diseases now have vaccines that mean outbreaks to these pathogens occur much less frequently (for example: measles, chicken pox, influenza etc.)

Let’s go through some examples of infectious disease outbreaks: viruses!

The COVID-19 pandemic of 2019

• The COVID-19 pandemic was caused by a virus known as SARS-CoV2. It caused a global pandemic.

• A pandemic is defined as an infectious disease affecting multiple parts of the world simultaneously, i.e. multiple countries are experiencing an epidemic.

• There are now several vaccines that have helped to reduce the severity of this infection.

Norovirus Outbreak 2023

• Norovirus causes vomitting and diarrhoea (you might call this gastro).

• It is extremely stable in the environment and highly contagious.

• Ingesting norovirus causes disease.

• There is no vaccine available for norovirus.

What about bacteria!

Buruli ulcer

Image: Institut Pasteur

• Buruli ulcer is caused by a bacteria known as Mycobacterium ulcerans.

• The bacteria is slow growing and flesh eating and is endemic to Melbourne and surrounds.

• Endemic means that there is an ongoing transmission of a pathogen in that area.

• Mode of transmission not entirely understood but likely involves mosquitoes and possums.

Pause after showing image

Who studies outbreaks

• Public health departments in the government

  • How to improve the health of everyone

• Academic research at Research Institutes and Universities

  • Make new discoveries, such as developing new drugs or vaccines

• Doctors and health staff

  • Keep their patients safe

And many others too !

An epidemiologist is someone who studies how…

“to keep the public informed on different public health issues and offer solutions to keep communities safe. They perform studies on outbreaks, their causes, transmission and effect on the public, collating that information into accessible data and health recommendations.”

Source: What is epidemiology and what does an epidemiologist do?

Ask students if they’ve heard of this before

What do we need to know about outbreaks?

• What is the problem?

• What is the cause?

• What can we do to make the situation better?

You are a disease detective for today and you must help us solve a mysterious outbreak!

The situation

Unknown illness causes students to fall ill at local school.

Local authorities are unsure of the cause of this outbreak, but it appears to be linked to a recent excursion to Melbourne CBD

The excursion

A few days earlier, students at a local school went into Melbourne to see an exhibition at the National Gallery of Victoria.

🚌 At 8am, they took the bus to the NGV.

🎨 They were in the gallery from 9am - 12pm.

🍲 They had lunch from 12pm - 2pm at three different locations.

🧑‍🎨 They went back to the gallery for another exhibition between 2 -4 pm.

🏡 At 4pm they were back on the bus to go home.

You are here to investigate…

And you have been given some information including:

• Name

• Class

• Age

• Sex

• Height

• Weight

• Was the student feeling sick

• What were their symptoms

And we are going to do this in R !

What does RStudio look like?

• Now to our computers
• You might be wondering what R looks like
• The top left is our R code file where we’ll be writing our code
• The top right is our environment pane which shows us R objects
• The bottom left pane shows us outputs of our code
• And bottom right currently shows us the folder that we’re working

Now get everyone to look at their computer and open screen to code with them

R can be our notebook

We can save important information for later.

We can use <- or =

my_name <- "Joy"
paste("My name is", my_name)

[1] "My name is Joy"

several_names <- c("Joy", "Anger", "Sadness")
paste("My name is", several_names)

[1] "My name is Joy"     "My name is Anger"   "My name is Sadness"

Your turn !

The same goes for numbers

a_number <- 137
another_number <- 120

a_number * another_number

[1] 16440

a_number / another_number

[1] 1.141667

(2*a_number) + another_number

[1] 394

Try saving a number or a word to a meaningful name - can you see where it goes?

• meaningful or descriptive word
• no spaces
• computer friendly to use underscore if you want to write multiple words
• try not to spend toooo long on this: max 2 mins

Now to solve an outbreak !

Where to start?

First we need to load in the data and tools to help us.

One of the great things about R is that there are lots of pre-installed commands that we can access or ones that others help to create and we can access. These are our “packages” that we access with the library() command.

We’ve set up a pre-written files to load this in for us!

source(here::here("setup", "setup.R"))
source(here::here("setup", "map.R"))

Shows them what’s on their screen and that we’ve done this for them already Show them how to run these commands by highlighting the ENTIRE code – easier Again try not to spend too long on this

What is outbreak_data?

💡The number one step in data analysis is looking at the data

Potentially highlight analogies here like reading a full recipe and seeing what ingredients you have or need to buy

For this, we are going to use our first pre-installed command called colnames().

colnames(outbreak_data)

 [1] "firstname"           "class"               "height"             
 [4] "weight"              "age"                 "sex"                
 [7] "temperature"         "sickness"            "shortness_of_breath"
[10] "chills"              "palpitation"         "bloody_stools"      
[13] "pain_chest"          "abdominal_cramps"    "dizziness"          
[16] "nausea"              "vomiting"            "vertigo"            
[19] "cough"               "fever"               "diarrhea"           
[22] "constipation"        "headache"            "pain_abdominal"     
[25] "throat_sore"         "muscle_pains"       

❓Can you find outbreak_data in the environment pane?

🧐 What happens if you click on it there?

Again try not to spend too long on this

Different kinds of data

In outbreak_data we have several kinds of information and it’s important to identify what these are !

• How would you describe what data we have in Firstname?

• How is this different to Weight ?

• What are some terms we could use to define these ?

• Again be quick
• Answers: character, numeric…

Visualising data

The types of data we have to work with will determine how it is best to visualise !

A scatter plot

Notice both the x and the y-axis are things we can count or measure.

A boxplot

Notice the y-axis are things we count or measure but the x-axis are specific groups.

Let’s plot our outbreak_data

We will be using a function called ggplot(). This is a fantastic tool we can use to make amazing graphs !

Like any graph we need to decide:

1. What do we want on the x-axis?
2. What do we want on the y-axis?
3. What kind of graph do we want to make?

🧑‍🎨other ideas like colour, design, groups … we’ll get to this later !

How to ask ggplot to plot outbreak_data

aes is referring to aesthetics, these are all the things we are going to customise.

  ggplot(data = outbreak_data,
         aes(x = class, y = height))

Next we choose what kind of graph we need

We can use geom_point() or geom_boxplot()

Let’s first try geom_point().

We use the + sign to add it to our first ggplot command

  ggplot(data = outbreak_data,
         aes(x = class, y = height))+
  geom_point()

Where would we put geom_boxplot() ?

We can use geom_point() or geom_boxplot()

Now swap height for temperature on the y-axis

  ggplot(data = outbreak_data,
         aes(x = class, y = temperature))+
  geom_point()

Let’s customise !

Can you change size, shape and colour…what kinds of plots can you make ?

  ggplot(data = outbreak_data,
         aes(x = class, y = temperature))+
  geom_point(size = 4, shape = 5, colour = "blue")

Let’s customise !

Let’s customise !

If you’re feeling stuck, try this:

scatterplot(outbreak_data, 
            x = class, y = temperature, 
            fill_col = "violet", 
            title = "My scatterplot")

Can you make a boxplot with class on the x-axis and temperature on the y-axis

max. 5 minutes

A boxplot summarising temperature in the four classes

  ggplot(data = outbreak_data,
         aes(x = class, y = temperature))+
  geom_boxplot()

A boxplot summarising temperature in the four classes

If you’re feeling stuck, try this:

boxplot(outbreak_data, 
        x = class, y = temperature,
        fill_col = "gold", 
        title = "My boxplot")

What does a boxplot tell us ?

• Provides us with an informative summary of the data in our groups
  • The center line gives the median
  • the box below is 25% of the data below the median and the box below is 25% of the data above

🧐 If we don’t know what R is doing, we can always get help !

?geom_boxplot()

How can we use fill to learn more ?

The fill can be specified as aesthetics like the x and y axis.

Can you specify the fill with our Sickness information about each student ?

  ggplot(data = outbreak_data,
         aes(x = class, y = temperature, fill = sickness))+
  geom_boxplot()

🧐 What does temperature tell us about how the students were feeling?

• Were the temperatures the same for students who were sick and those who weren’t

• Where the sick students in all classes

• Why do people get a temperature when they get sick?

= Max: 5 minutes

Time to give our plot a title

  ggplot(data = outbreak_data,
         aes(x = class, y = temperature, fill = sickness))+
  geom_boxplot()+
  labs(title = "Make a title here !")

Make it something informative that describes the plot you’ve just made and why people would be interested to look at it

Let’s calculate some summaries

This symbol %>% is used to send our data into another command such as summarise()

As the name would suggest, we can generate summaries, such as calculating the median

outbreak_data %>% 
  summarise(median = median(temperature))

# A tibble: 1 × 1
  median
   <dbl>
1     37

We can also use group_by() with multiple groups

Just like our boxplot, we want to know the median temperature of the students, grouped by their Class and Sickness status

👏Try to do this by adding group_by(class, sickness)

outbreak_data %>% 
  group_by(class, sickness) %>% 
  summarise(median = median(temperature))

How can we relate this to the boxplot we made earlier ?

We can also use group_by() with multiple groups

# A tibble: 8 × 3
# Groups:   class [4]
  class sickness median
  <chr> <chr>     <dbl>
1 A     no         36.5
2 A     yes        39  
3 B     no         37  
4 B     yes        39  
5 C     no         37  
6 C     yes        39  
7 D     no         37  
8 D     yes        39  

Let’s focus on the student’s symptoms

We are going to use a new type of graph to do this, it’s called a bar plot.

For this we use geom_bar()

ggplot(data = outbreak_data,
         aes(x =  sickness))+
  geom_bar()

How many students were and were not sick ?

As well as the bar plot, we can also use the count() command

outbreak_data %>% 
  count(sickness)

# A tibble: 2 × 2
  sickness     n
  <chr>    <int>
1 no          18
2 yes         11

What is making the students sick ?

Sickness summarises all of the symptoms. Now it’s time to get specific!

Can you make some other bar plots to see what the students were sick with?

💡 Remember we can use colnames(outbreak_data) to find out what information we have in outbreak_data!

🤩 You might also like to try customising your plot. Remember we can change the colour by specifying fill = within the geom_bar() command. You might also like to try aes(fill = ).

Shortness of breath

  ggplot(data = outbreak_data,
         aes(x = shortness_of_breath))+
  geom_bar(aes(fill = shortness_of_breath))

🧐 What have you found? What symptoms are the most common ?

max. 5 minutes

Re-organising our symptoms data

Data can be formatted in different ways.

For example, prioritise having one row per individual with lots of columns.

This is like our outbreak_data.

Or sometimes we can change the format and so instead we have a long list of one observation per row.

This is what we have in our symptoms data.

For the next few plots we are going to use symptoms and not outbreak_data, however they record exactly the same thing.

Bringing all the symptoms together

ggplot(data = symptoms,
       aes(y = symptom, fill = status))+
  geom_bar()

Bringing all the symptoms together

• This doesn’t necessarily help us to decide what is making the students sick because they are not separated as to whether they were sick or not

• We really want to separate those that have a YES in Sickness

We can filter this list to just those that are sick !

This needs the filter() command

symptoms_sick <- symptoms %>% 
  filter(sickness=="yes")

Now, what symptoms do the sick students have ?

ggplot(data = symptoms_sick,
       aes(y = symptom, fill = status))+
  geom_bar()+
  labs(title = "What symptoms do the sick students have ?")

Now, what symptoms do the sick students have ?

barplot(symptoms_sick, 
        x = symptom, 
        colour_by = status, 
        horizontal = TRUE, 
        title = "What symptoms do the sick students have ?")

Interpreting our data

What does this graph tell us?

• What are the symptoms of those that are sick?

• What might be some causes of illness that give these symptoms?

Likely cause…

ask if they have any guesses

Salmonella

What is salmonella?

• A bacterial infection

• Commonly from contaminated food

• Examples…undercooked chicken, raw egg, insufficient cleaning

• You can start to feel sick within 6 to 36 hours after eating

But how did the students get sick ?

Challenge: Where did the students get sick?

🎨 At the gallery?

🚌 On the bus?

🍲 At lunch?

WAIT ! We have some more information

🍲 We have just been sent some information about where the student’s ate lunch !

They ate at three different restaurants !

max. 5 minutes

Plot where the students ate their lunch…

ggmap(melb_map)+
   geom_point(data=map_data, 
              aes(x =long, y = lat, colour = sickness), 
              size = 12, shape = 1)

This was based on a real life outbreak in Melbourne !

ABC New 2015

Today we have learnt…

🦠 What are pathogens and what are outbreaks

🌎 Who are epidemiologists

🔢 What is R

📊 How to make plots with ggplot

📈 How to organise and arrange complex data

🤠 What’s something that you’ve learnt today?

Do you want to learn more?

Our online book of today’s lesson

Scan the QR code on our poster!

RStudio Education for Beginners

R for Data Science

The Epidemiologist R Handbook

R You With Me

R-Graph Gallery

And thank you to…

It Takes a Spark !

and to all of you for helping solve the outbreak !