Microbial Antibiotic Resistance
MAT 259, 2021
Ashley Bruce
Concept
As a student with a background in Biology, for my project, I wanted to incorporate a project that had something
to do with microbial data. After browsing through experiments, I found one that I thought would be both interesting,
and visualize well: E.coli resistance to differing antibitiocs and phages.
Query
Data was obtained through Dryad.org. The specific study that I obtained the data from is titled: "Associations between
sensitivity to antibiotics, disinfectants, and heavy metals in natural, clinical and laboratory isolates of
Escherichia coli". A link to this article, the page where the raw data can be downloaded, and the raw data will be included
in the zip file for this project.
If one downloads all the data, it can be seen that there are many csv files related to this study. After looking through all of them, I found one that would work best for the purpose of this project. This file is called "Raw_Data.csv". Aside from the deletion of a few columns, the rest of the data processing for this project was done through Processing.
If one downloads all the data, it can be seen that there are many csv files related to this study. After looking through all of them, I found one that would work best for the purpose of this project. This file is called "Raw_Data.csv". Aside from the deletion of a few columns, the rest of the data processing for this project was done through Processing.
Preliminary sketches
Because this project has so many variables to work with, I was trying to think of the best way to visualize the data that made it
clear and easily understandable. The result I came up with is below.
The reasoning for my sketch is outlined in the process section of this page.
The reasoning for my sketch is outlined in the process section of this page.
Process
As the data from this experiment in Biological in nature, there are a few key terms that I feel need explaining before the sketch
and the final visualization make sense.
The project where I pull the data from does their research on the E.coli bacteria. There are many different strains of E.coli. Having a different strain means that the genetic material of one strain differs in some key way from another strain, making the two strains similar enough to still be considered the same species, but different enough to be different strains. A good analogy to this would be to think of E.coli as "fruits". Two different strains of fruit could be an Apple and Banana. They're different, but still within the group of fruit.
Aabout 100 different strains of E.coli are being tested on in this study. It is important to differentiate the different strains because different strains will potentially have differing resistance to treatments. This is the reason I wanted to be able to separate the strains from one another.
To test the resistance of E.coli to different things, the strains are exposed to one of 26 treatments at a time. Note, that for each strain of E.coli, it is being subject to all the treatments, just one at a time.
Furthermore, each treatment is given to a strain at differing concentrations. I will write an example below to help make this more clear.
Example:
Consider a strain of E.coli -- Strain 1.
Strain 1 will be subject to 2 treatments: Amoxicillin and Erythromycin
Each of these treatments will be offered in 2 concentrations: 1ug and 5ug
This means that in order to consider all the possible combinations the experiments would happen like so:
Strain 1 with Amoxicillin at 1ug, Strain 1 with Amoxicillin at 5ug, Strain 1 with Erythromycin at 1ug, Strain 1 with Erythromycin at 5 ug.
If we introduce Strain 2 now, we have to repeat the above just now with Strain 2.
After each of these treatments is carried out, we need some way to measure the amount of bacterial still present in the sample. This is done via spectroscopy. More specifically, spectroscopy done at 600nm. By looking at the optical density of a sample measured at a wavelength of 600nm, we can estimate the amount of bacteria in a sample. This experiment measured the OD at 600 for all the samples, and is how results are quantified.
The project where I pull the data from does their research on the E.coli bacteria. There are many different strains of E.coli. Having a different strain means that the genetic material of one strain differs in some key way from another strain, making the two strains similar enough to still be considered the same species, but different enough to be different strains. A good analogy to this would be to think of E.coli as "fruits". Two different strains of fruit could be an Apple and Banana. They're different, but still within the group of fruit.
Aabout 100 different strains of E.coli are being tested on in this study. It is important to differentiate the different strains because different strains will potentially have differing resistance to treatments. This is the reason I wanted to be able to separate the strains from one another.
To test the resistance of E.coli to different things, the strains are exposed to one of 26 treatments at a time. Note, that for each strain of E.coli, it is being subject to all the treatments, just one at a time.
Furthermore, each treatment is given to a strain at differing concentrations. I will write an example below to help make this more clear.
Example:
Consider a strain of E.coli -- Strain 1.
Strain 1 will be subject to 2 treatments: Amoxicillin and Erythromycin
Each of these treatments will be offered in 2 concentrations: 1ug and 5ug
This means that in order to consider all the possible combinations the experiments would happen like so:
Strain 1 with Amoxicillin at 1ug, Strain 1 with Amoxicillin at 5ug, Strain 1 with Erythromycin at 1ug, Strain 1 with Erythromycin at 5 ug.
If we introduce Strain 2 now, we have to repeat the above just now with Strain 2.
After each of these treatments is carried out, we need some way to measure the amount of bacterial still present in the sample. This is done via spectroscopy. More specifically, spectroscopy done at 600nm. By looking at the optical density of a sample measured at a wavelength of 600nm, we can estimate the amount of bacteria in a sample. This experiment measured the OD at 600 for all the samples, and is how results are quantified.
Final result
The results can be seen in the images below.
Here is the plain data for the first strain.
Using the control menu on the left, the user can play around with how the data is seen. In this picture, antibiotic treatments are highlighted in green, while treatment with phages is darkened as blue.
To also help differentiate the treatments from one another, the user can make each treatment a different color. The fill is is taken from the max line.
Furthermore, instead of looking at the max OD readings, the user can select to look at the average line to better see the effectiveness of the treatment.
Seeing the results from this experiment brought about a few things that surprised me. As to be expected, for the most part, as the concentration increased, the OD decreased, showing a decrease in the number of E.coli in the sample. This meant that the treatment was effective. In the cases where this is not observed, it means that the strain is conferring some sort of resistance to that treatment. Surprising to me though, was that for the phages, treatment was only done in a few concentrations, and the lack of decrease in OD reading means that the strains seemed to have some resistance to the phages.
Here is the plain data for the first strain.
Using the control menu on the left, the user can play around with how the data is seen. In this picture, antibiotic treatments are highlighted in green, while treatment with phages is darkened as blue.
To also help differentiate the treatments from one another, the user can make each treatment a different color. The fill is is taken from the max line.
Furthermore, instead of looking at the max OD readings, the user can select to look at the average line to better see the effectiveness of the treatment.
Seeing the results from this experiment brought about a few things that surprised me. As to be expected, for the most part, as the concentration increased, the OD decreased, showing a decrease in the number of E.coli in the sample. This meant that the treatment was effective. In the cases where this is not observed, it means that the strain is conferring some sort of resistance to that treatment. Surprising to me though, was that for the phages, treatment was only done in a few concentrations, and the lack of decrease in OD reading means that the strains seemed to have some resistance to the phages.
Code