BME100 f2014:Group15 L6

From OpenWetWare
Jump to navigationJump to search
Owwnotebook icon.png BME 100 Fall 2014 Home
Lab Write-Up 1 | Lab Write-Up 2 | Lab Write-Up 3
Lab Write-Up 4 | Lab Write-Up 5 | Lab Write-Up 6
Course Logistics For Instructors
Wiki Editing Help
BME494 Asu logo.png


Name: Michael Catchings
Name: Jessica Fong
Name: Ben Heywood
Name: Norihan Elsharawy
Name: Destiny Vidaure
Name: Logan Migliorino


Bayesian Statistics

Overview of the Original Diagnosis System

In this investigation, the BME 100 class utilized PCR analysis to determine the existence of the disease-associated SNP in multiple people. In order to run the tests and acquire accurate data, 34 groups of six students ran 186 PCR tests and diagnosed 68 patients with the disease-associated SNP. Each individual group ran their own PCR tests on the DNA of two different patients to determine if they were negative or positive for the SNP. The results were compiled into one large document, and Bayesian calculations to analyze the data and determine our probability with our results. In order to prevent error within our testing, Dr. Haynes used her significantly more precise PCR machine and analyzed the three replicates of DNA per patient for both patients for the BME 100 class. When analyzing the fluorescence of SYBR Green fluorescent, we ensured that the fluorimeter and camera system (for later analysis in ImageJ) were isolated from surrounding light via the dark box. We also analyzed two control samples of DNA, one that was negative for the SNP, and one that was positive for the SNP, in order to ensure our SYBR green would detect the SNP in our experimental samples. Positive results would glow green when exposed to UV light, negative results would remain clear. For the ImageJ image analysis program, which was used to analyze the concentration of SYBR green fluorescence in the experimental samples, the split channel tool was applied, and only green colored images were analyzed. The other displayed colors of blue and red of the RGB color scheme were discarded in the analysis. Each calibration trial used three images for analysis, and the mean average taken and used for our data.

The final data used for analysis were of 186 tests with 68 diagnosed patients. Of this data, it was given that 30 patients were positive and 24 were negative. The compiled results, however, showed that 23 were positive and 45 were negative for the disease-associated SNP. Some of the data was also inconclusive, demonstrating errors in experimentation or PCR testing, and was discarded. Other data that was blank, due to errors in it's analysis, and was discarded. We found that PCR is, by itself, not a sure way to effectively test for disease associated SNPs, and should be used in conjunction with other tests to ensure accuracy.

What Bayes Statistics Imply about This Diagnostic Approach

Calculation 1
A Frequency of Cancer-Positive Conclusions 0.48
B Frequency of Positive PCR Relations 0.49
P(B│A) Probability of Positive PCR Relations Given Cancer-Positive Conclusions 0.89
P(A│B) Probability of Cancer-Positive Conclusions Given Positive PCR Reaction 87%

Calculation 2
A Frequency of Cancer-Negative Conclusions 0.39
B Frequency of Negative PCR Relations 0.39
P(B│A) Probability of Negative PCR Relations Given Cancer-Negative Conclusions 0.77
P(A│B) Probability of Cancer-Negative Conclusions Given Negative PCR Reaction 77%

Calculation 3
A Frequency of Cancer "Yes" Diagnosis 0.34
B Frequency of Positive DNA Test Conclusion 0.48
P(B│A) Probability of Positive DNA Test Conclusion Given Cancer "Yes" Diagnosis 0.43
P(A│B) Probability of Cancer "Yes" Diagnosis Given Positive DNA Test Conclusion 30%

Calculation 4
A Frequency of Cancer "No" Diagnosis 0.66
B Frequency of Negative DNA Test Conclusion 0.39
P(B│A) Probability of Negative DNA Test Conclusion Given Cancer "No" Diagnosis 0.27
P(A│B) Probability of Cancer "No" Diagnosis Given Negative DNA Test Conclusion 46%

In the first and second Bayesian calculations, there was an emphasis on figuring out if the PCR concluded a positive or negative conclusion in the patient’s DNA. A positive conclusion would indicate that the PCR found a disease-associated SNP in the sample, likewise, a negative conclusion would indicate the absence of the disease-associated SNP in the sample. As seen above, the values of both the positive conclusion and negative conclusion are close to 100%, but still have around a 13-23% error. This error value indicates that the PCR can be wrong with its readings, as the samples inputted into the PCR were known to be either possessing the disease or not. Looking at the real world application of PCR testing, the data in this lab shows that other methods for detecting disease-associated SNP might be favored because of the lack of reliability in PCR method. Some possible sources of error in this lab included several faulty PCRs, possible inaccurate micro-pipetting, and the camera resolution when photographing SYBR green.

The third and fourth Bayesian calculations were used to determine a “yes” or “no” diagnosis of disease-associated SNP from the positive and negative conclusions calculated in the first two calculations. A “yes” reading shows that the patient does have the cancerous disease, and a “no” reading shows that the patient does not have the disease. In the table of calculations, these two values are displayed very small compared to the 100% standard. This means that if a sample of DNA is concluded positive for the SNP by the PCR, there is a small chance that the patient actual has the disease, and vice versa for a negative conclusion. This error once again adds to the lack of reliability in PCR testing.

Computer-Aided Design


TinkerCAD is a free online design tool that allows users to create digital designs. TinkerCAD has users begin with guided lessons to walk them through the design process and tools and features available in TinkerCAD. After going through the lessons and some tutorials to learn how to use TinkerCAD, we used TinkerCAD to create a 3D model of our design of how we would improve the PCR machine.

Our Design

The new design for the Open PCR machine includes the addition of clamps on the outside of the machine and an improved tray to hold the test tubes. The clamps would be added to the top of the PCR in order to more effectively hold down the lid and prevent heat loss. The tray that holds the test tubes during the PCR process would be made of a metal that is a better conductor of heat so that it can warm up and cool off much more efficiently. The area that the tray is sitting in would have to be a good insulator to keep the temperature constant which is why the physical tray needs to be an effective conductor. These changes are necessary because the DNA needs optimal conditions to satisfy the temperature requirement of the DNA Polymerase for it to make as many copies as possible.

Feature 2: Consumables Kit

The kit will include two long vertical compartments and three smaller horizontal compartments alongside each other. The two long vertical compartments will have a divider to separate the micropipette from it tubes. The liquid reagents, such as the PCR mix and primers, will be placed in the two of the horizontal compartments and the plastic tubes will be placed in the last horizontal compartment. The kit will have lid to cover the material inside the kit.

\Description of image

Feature 3: Hardware - PCR Machine & Fluorimeter

The new system will be very similar to the devices now. The only difference for the PCR machine will be the addition and revitalization of some parts, including the clamps on the exterior portion of the lid and the newly designed tray. For the fluorimeter, the outer flap that is suppose to be closed before the picture is taken will have a small area cut out so that a person's arm can fit underneath, but the lid will still close completely. This should reduce any error that occurs because of an additional light source.

Description of image