BME100 f2014:Group18 L6

From OpenWetWare
Jump to: navigation, 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: Norah Alkhamis
Name: Jesus Calderon
Name: Kevin Couch
Name: Jordan Kariniemi
Name: Scott Slade
Name: Rachel Tomlinson


Bayesian Statistics

Overview of the Original Diagnosis System
To begin, each group in the BME 100 class was given two patients, and instructed to test these patients for the disease-associated SNP. The total laboratory consisted of 34 teams (consisting of 6 students each team) with two patients assigned to each team. All together, there were 68 patients that were being tested for the disease. Originally, while testing the droplets of DNA with the fluorimeter, each testing sample had three pictures taken of it under the light of the fluorimeter. These images were then grouped, analyzed, and then the data was averaged in hopes of finding the most accurate data for each specific sample. Each patient had three replicants of their DNA tested for the disease. The mean of this data was then taken to continuously potentially limit any error. The BME 100 class data was comprised into a master spreadsheet that included all of the teams' work/data. The data across the groups varied. Some groups' data had successfully concluded their sample data, while others received inconclusive results, or left their information blank. By observing the class results, it was concluded that there were 30 total positive results, and 24 total negative results with 8 inconclusive results, and 6 blank results. Possible sources of error may be due to the multitude of teams in the whole BME 100 class. Because there were 34 teams, this suggested that there were 34 different ways to perform the lab, meaning the data could have been received in a different manner each time.

What Bayes Statistics Imply about This Diagnostic Approach

Overall, calculations 1 and 2 imply that the individual PCR replicates were very reliable in concluding that a person has the disease SNP or doesn't. The Bayes value for calculation 1 was very close to 1.00 (100%), and the Bayes value for calculation 2 was also close but not as close as calculation 1. Some possible errors that could have affected the Bayes value in a negative way include definite human error in pipetting the substances correctly in technique and amount of solution. Furthermore, during the fluorimetry, there could have been major errors in blocking out the light correctly thus causing the focusing adjustment of the camera to vary, which would have ultimately affected the images captured. Another source of human error could have been in preparing the solutions correctly and in putting them into the PCR machine correctly. Unlike the results for calculations 1 and 2, the results for calculations 3 and 4 suggest that the PCR is not very reliable in predicting the development of the disease (diagnosis) given a positive or negative test. Calculation 3 had a low probability of predicting the development of the disease correctly while calculation 4 had a higher possibility of predicting the development of the disease correctly with close to a 50% probability of correctly predicting the results. Both Bayes value for calculations 3 and 4 were very small and not close to 1.00 (100%).

Description of image
Description of image
Description of image
Description of image

Computer-Aided Design


Using the TinkerCAD program, our group was able to refer to and analyze the standard open PCR machine in 3-D to develop an idea for a way it could be improved. With each idea of improvement, the implications of the change were discussed. For example, we considered if the PCR machine was designed to be able to hold more samples, the process would take longer, and most likely need another heating/cooling element, thus causing a need for a larger powered circuit board. In the end, we decided as a group that we should add another cooling element to the existing PCR design. This would not significantly change the design of the PCR machine but it would allow for the final PCR reaction to be reached quicker and more efficiently.

Our Design

Description of image

We added a cooling element to the Open PCR machine. The process for the PCR reaction requires a cooling stage, and this added element allows for the test tube samples to reach the cool temperature in shorter time and more efficiently. Ultimately, it is reducing the total time required for the PCR reaction.

Feature 2: Consumables Kit

We will hold the SYBR Green 1 in light blocking packaging which would replace the original tin foil packaging used in the lab. The packaging would also be water proof to prevent cross contamination. The buffers needed to be sterile to assure the cleanliness and accuracy of the lab. The packaging for the micropipettor tips should be cut in half so that they could all be used in one sitting for the lab to allow for the most sterile process.

Feature 3: Hardware - PCR Machine & Fluorimeter

For the fluorimeter, it would be best to include an adjustable camera holder to allow for varying heights to help the camera flush with the side of the PCR machine. We would also change the folding lid of the fluorimeter to allow complete darkness, but still the ability to capture a clear image. The lack of these two features in the existing fluorimeter design, that we experimented with, caused several problems. Without the adjustable camera holder, there were changes in cameras to test who's worked better and provided a clearer image. Eventually, the entire fluorimeter set up had to be redone. Additionally, lowering the flap with the camera on a timer caused a lot of glare and problems because of the camera adjusting to the light. In order to get a clear picture, the lab had to be slightly compromised by allowing the flap to be open partially to at least allow some light to shine into the fluorimeter.