BME 100 Lab students tested patients for a disease-associated SNP by using open PCR and Fluorimetry. Sixteen teams each with five or six members diagnosed a total of thirty-two patients total. To prevent error, there were three samples of DNA per each patient as well as a positive control and a negative control used as a standard for comparison. The micropipette tips were discarded and replaced after each use to avoid cross-contamination. Each drop during Fluorimetry had three pictures taken to ensure consistent data for the ImageJ calculations. The pictures were all taken with the same phone with the same settings to also ensure consistency when analyzing the data with ImageJ. The phone was supposed to be the same distance away from each drop on the fluorimeter (4 cm) each time a picture was taken. Out of thirty patients, there were only twenty-eight conclusions made (one group never submitted their data). Out of the twenty-eight conclusions, seven came back positive (meaning patient has SNP), nineteen came back negative (patient not having SNP), and two came back inconclusive. There were a couple of factors that may have affected our data though. One was the screw down lid of the PCR machine. There was difficulty with screwing it down. If it was not locked tight enough, it would have affected the temperature of the system because it could have allowed heat to escape. This could have prevented the DNA from replicating itself enough times and therefore the concentration of the SNP would be different. Another problem was with the phone holder. It was hard to find get the phone the desired distance from the fluorimeter since it was not attached to the fluorimeter. This meant that not all of our drops were the same size in the pictures. Although this did not seem to affect our ability to make a conclusion, it probably affected the calculations from ImageJ.
What Bayes Statistics Imply about This Diagnostic Approach
When it came to replicating the conclusion that a person has the disease SNP or not, the PCR, overall, was pretty reliable. For replicating a positive conclusion, the PCR had a Bayes value in the middle between .5 and 1.0. While it’s not super reliable at replicating the conclusion for if a person has the disease, it is still pretty reliable and, most likely, could be used to test for it. For replicating a negative conclusion, the PCR was much more reliable, having a Bayes value much closer to 1.0 / 100%. In this respect, the PCR is very reliable. Overall, the Bayes alue show that the individual PCR replicates for concluding that a person has the disease SNP or not is pretty reliable, and can be used to, for the most part, test for whether or not the SNP is present.
When it came to predicting if a person will develop the disease, the reliability of PCR, according to Bayes calculations/ values, was closer to .5 / 50%. This indicates the PCR is not super reliable, and is not the best indicator/ predictor for if the person will develop the disease. However, when it came to predicting if the person would NOT develop the disease, the Bayes value was much closer to 1.0 / 100%. This indicates that PCR is reliable at predicting when someone will not develop the disease. While it is reliable for predicting the negative, the PCR is not reliable for predicting the positive, which, depending on the disease the PCR / fluorimeter is testing for, could prove to be fatal. According to this data, the PCR would not be a reliable way to test for the development of the disease.
During the fluorimeter steps of taking pictures of the individual sample droplets of the experiment, our group, as well as several others we have talked to, concluded that it was very difficult to keep the phone at a consistent distance from the droplet to ensure that all were the same sizes. This caused the droplets to be different sizes and made the analyzing in imageJ hard and not as reliable.
During the pipetting phase of the PCR material to prepare it for the fluorimeter it would have been easy to cross contaminate the PCR products with each other if someone was not being cautious in the transfer. This would have cross-contaminated the PCR products with each other and could have led to a false positive or false negative.
This would tie back into the importance of labeling the containers with the right patient number vs the controls. If any of these were mislabeled or illegible this would/could cause the result to be wrong for a certain patient or control. This is why making sure to label every aspect ant tube of the lab is vital, because without the label it is impossible to set standards or test the patient's DNA against the disease SNP.
What is the probability that a patient will get a positive final test conclusion, given a positive PCR reaction?
Variable
Description
Numerical Value
A
Positive Conclusion
7, P(A)=0.25
B
Positive PCR Reaction
23, P(B)=0.274
A)
Probability of a positive PCR reaction given a positive conclusion
0.8696
B)
Probability of a positive final test conclusion given a positive PCR reaction
0.79395
What is the probability that a patient will get a negative final test conclusion, given a negative diagnostic signal?
Variable
Description
Numerical Value
A
Negative Conclusion
19, P(A)=0.67857
B
Negative PCR Reaction
53, P(B)=0.631
A)
Probability of a negative PCR reaction given a negative conclusion
0.906
B)
Probability of a negative conclusion given a negative PCR reaction
0.9743
What is the probability that a patient will develop a disease, given a positive final test calculation?
Variable
Description
Numerical Value
A
Will develop a disease
8, P(A)=0.267
B
Positive Conclusion
7, P(B)=0.25
A)
Probability of a positive conclusion given that the patient will develop a disease
0.571
B)
Probability that a patient will develop a disease given a positive conclusion
0.6098
What is the probability that a patient will not develop the disease, given the negative final test conclusion?
Variable
Description
Numerical Value
A
Will not develop disease
22, P(A)=0.733
B
Negative conclusion
19, P(B)=0.67857
A)
Probability of a negative conclusion given they will not develop disease
0.842
B)
Probability that a patient will not develop the disease, given a negative conclusion
0.90953
Which calculation describes the sensitivity of the system regarding the ability to detect the disease SNP?
0.79395
Which calculation describes the sensitivity of the system regarding the ability to predict the disease?
0.9529
Which calculation describes the specificity of the system regarding the ability to detect the disease SNP?
0.6098
Which calculation describes the specificity of the system regarding the ability to predict the disease?
0.90953
Intro to Computer-Aided Design
3D Modeling
Our team used SolidWorks to make the new design. While it was a bit of a hassle to use at first, after some time it was another walk in the park, and a fun one at that. Considering users are not allowed to work on multiple parts at once it was a chore to work in more than one plane. This was especially true when having to make a part that had only rounded sides or making a part with many components. The reason this was an issue was that a flat plane was required to make a new sketch for a part, and it was either difficult to work on a small plane or there was no flat areas to work with. Aside from the multi-dimensional sketching, working with SolidWorks was fairly easy. It was a little less intuitive or UI friendly than other CAD programs used, but that does not discount it in any way. Making assemblies is a breeze considering it is only dragging and dropping files, and having multiple parts open at a time made the hassle of singular part drafting almost nonexistent. As a whole, SolidsWorks is going to make a lovely tool in any engineer’s toolbelt.
Our Design
We chose this design because it will keep the camera at a set distance and position from the fluorimeter. The original design lacked any way to lock the stand to the fluorimeter and the phone stand was susceptible to being mmoved which affected the photos. Because the stand was not locked in place, it resulted in inconsistency in the size and orientation of the droplets in ImageJ. Our device allows the position to be locked and adjusted to ensure consistent data and pictures.
Feature 1: Consumables
The consumables we would include in the kit are consumables required for the fluorimeter. This means we would include the fluorimeter’s box, the parts for our new stand, three glass slides, a light bulb (for the fluorimeter) as well as 2000 microliters of the SYBR Green solution and 5000 microliters of the buffer solution.
Feature 2: Hardware - PCR Machine & Fluorimeter
The open PCR and the fluorimeter will be packaged separately. The PCR machine will come fully assembled along with its consumables. The fluorimeter and our phone stand will need to be assembled and will come along with its consumables. The The open PCR machine and the fluorimeter will be used in the same exact way as before. Our design is simply an addition to the fluorimeter that only changes the picture outcome (to make them all very similar in size / orientation) but does not affect how the machines work. Because it is a simple addition, no actual change to the Open PCR or fluorimeter or how they work / their purpose in this lab was involved, resulting in no change to how they will be used. The Open PCR is still used to amplify the DNA segments while the fluorimeter is still used to capture the concentration (through pictures) using luminescence.
The way our group has decided to make additional changes to the fluorimeter set up is by altering the stand used for our camera. The stand has been altered to allow for rubber arms to be attached to the base. The reason for this is that this allows for our stand to remain stationary between different trials, which was a problem experienced during the lab. The arms have extra length at where they connect to the stand to allow for the arms to come closer together should they need to. The rubber also has multiple uses which is why we opted for that. First off, the rubber can bend outward slightly without moving the stand which allows for the arms to hold onto larger items. Second off, the flat rubber ends provide enough friction to prevent any slipping.
BME 100 Fall 2017
