BME100 f2016:Group8 W8AM L6

From OpenWetWare
Jump to: navigation, search
Owwnotebook icon.png BME 100 Fall 2016 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: Jasmine Davis
Name: Dhantin Kumar
Name: Blake Sentyrz
Name: Juan Pablo Robayo
Name: Andre Bordeleau



Bayesian Statistics

Overview of the Original Diagnosis System There were 34 patients that were diagnosed and were divided among 17 groups, with each group consisting of around five people; each group got two patients. Each group made three replicates for each patient, so in the end there were 6 replicates per group. Each group was responsible for testing the samples of DNA from their two patients and calculating the DNA concentration in ug/mL of the DNA samples to further investigate by using the OpenPCR machine that amplifies the DNA in the samples for further investigation in the fluorimeter. Each group was given a positive and negative control in order to compare the patient to and decide whether they tested positive or negative for the disease. A calibration curve was created with the known sample concentrations, and the patients' DNA concentrations were then compared to the curve and calculated based on the curve. When ImageJ was applied, there were three drop images for each patient in order to get accurate calculations. This was done using the fluorimeter, which took images of the sample droplets in SYBR Green 1 solution. This solution mix will glow to signify that the sample being tested has the marker for the DNA we are looking for in the lab. ImageJ was used to find the Mean Pixel Value, Area, and RAWINTDEN data. These results of all the patients were then compiled into a spreadsheet which was then used to solve for the Bayesian Statistics. A glowing image signified a positive sample and a non-glowing image represented the negative sample for the DNA marker that was being investigated in this lab. There were two groups who had blank data- groups 10 and 14, and therefore the class's final data consisted of 84 total PCR tests, with 38 positive PCR results, 40 negative PCR results, and 6 inconclusive. Because there were three images per patient, there were a total of 28 conclusions, with 13 positive conclusions and 15 negative conclusions. When compared to the total diagnoses- which were also 28- there were 12 patients that had the disease and 16 patients who did not have the disease. As a result, there were 9 patients that had the disease and tested positive, and 12 patients who did not have the disease and tested negative. Our group had issues in the calculations, specifically in calculating the concentrations of the positive and negative controls. And although both of our conclusions matched the doctor's diagnoses, one of our PCR results were deemed inconclusive.

What Bayes Statistics Imply about This Diagnostic Approach In general there was about a 50% (69%) chance of a patient getting a positive PCR result as well as a positive test conclusion. But within that probability, there was about a 90% chance of someone that had a positive PCR result having a positive test conclusion. The second calculation showed that there was around a 50% chance of someone receiving negative test results or a negative test conclusion. Within those numbers there was an 80% chance of someone having a negative PCR result and not having the disease, and a 90% probability of someone having the disease and getting a negative PCR result. Overall, the individual PCR replicates are reliable when done in large numbers, as was this experiment, as the experiment statistically shows us that the PCR replicates are reliable in determining if the SNP disease is present or not present in the patient's DNA.

The third calculation showed that there is about a 50% chance of someone being diagnosed with the disease, as well as receiving a positive PCR result. However the probability of having a positive test conclusion when the patient has the disease is less that 50% accurate, and the probability of someone being diagnosed with a final positive test is only 70%. The fourth calculation showed that for someone who does not have the disease, the probability of that patient receiving a negative test conclusion is only 50% accurate. However the probability that someone does not have the disease given by the fact that the final test result was negative is 80%. Overall, using the positive or negative test results is not reliable means of determining if the patient will develop the SNP disease in their DNA, as the results show only a 50% accuracy.

Possible Sources of Error

  1. Faulty Camera: The camera quality may have affected the overall resolution of the images that were produced, which would then effect the wuality of the images used in the ImageJ software, since the ImageJ software uses these images to determine the size of the droplet with the glow of the DNA marker. Forever, faulty camera setting may contribute to miscalculations.
  2. Light exposed in fluorimeter: SYBR Green I solution's fluorescence capabilities may have been affected by the light that streamed through the cracks of the fluorimeter box and when mixing the solutions, which would then reduce the effectiveness of the solution to glow in the dark. This would then affect the results since the SYBR green may or may not have glowed to its full capacity in the image, therefore making the image results faulty.
  3. Pipetting: Pipetting the samples into the specific tubes may have been an issue since without proper pipetting technique, the solution samples may have been compromised, and since this reaction occurs on such a small scale, such mistakes may radically change the results of the experiment by not allowing for proper DNA amplification.

Intro to Computer-Aided Design

3D Modeling
The software our group used to design our product was called TinkerCAD, and was relatively simple to use and work with compared to SolidWorks. This program allowed our team to view a product in the 3D realm, so that the designing of the product was more realistic as if we had the parts to build the machine with us. Being a 3d-spatial modeling software with preloaded designs of different parts, it was up to our team to use these parts to create our own design for our PCR machine. Each member of our team created our own designs for the machine, and after collaboration, our team compiled one final model of what we wanted our machine to look like. Ultimately, TinkerCAD was easier to work with than SolidWorks and was fairly simple to learn.

Our Design

Image View
Front View
Top View
Back View
Side View

Our design for this new PCR machine-- the machine we are dubbing "duoPCR"-- focuses on combining two aspects of the PCR process, the PCR thermal cycling and the fluorimeter, into one, easy to use machine. DuoPCR, which was named after the idea that this PCR machine has "dual" capabilities of a PCR machine and a fluorimeter, eliminates the need for a researcher to have two different machines to understand the data from the PCR machine. Our machine reduces the time spent working to understand the information gathered by the PCR machine because the built-in fluorimeter captures the pictures of the droplets of samples, and, working with the built-in ImageJ software in the touchscreen LCD monitor on the top, will display all the results of the sample fluorimeter. The bottom and the back of the PCR machine remain the same as the OpenPCr machine we used in lab. However, every other aspect of the PCR machine was redesigned. The thermal cycling box now sits on top of the box while all the components necessary for thermal regulation of the samples in the DNA amplification stage now lie on the bottom of the machine and in the legs of the machine, as seen in the images above. Moreover, the thermal cycling box is connected to the touchscreen LCD monitor, which allows for easy reading of the samples and the information. Underneath all of this is the fluorimeter, which has built in cameras that are attached to the monitor on top for easy download of information and has a pull-up, retractable door to minimize the amount of light that enters the fluorimeter. This way, there is no hassle in adjusting a separate camera phone, and the information that is captured by the built-in fluorimeter is easily displayed on the monitor and able to be easily downloaded onto a separate machine. Our model allows for a faster, more efficient manner of PCR process by combining two aspects of the PCR process into one. It will save time by cutting down the hassle spent on transferring samples between machines, and will help the researcher by reducing the amount of work that he/she has to do.

Feature 1: Consumables

A problem that arose in this lab was the mismanagement of samples because of faulty labels. To combat this issue, our product will incorporate a rack in the thermal cycler box on top of the machine with labels on the racks so that the samples will not be mis-labeled. However, no further consumables will be packaged with the machine as the machine is built for all standard consumables.

Feature 2: Hardware - PCR Machine & Fluorimeter

Our PCR machine, the duoPCR, aims to combine two aspects of the PCR process into one, easy to use machine, serving "dual" purposes, PCR thermal cycling and fluorimeter Our system will include the PCR thermal cycler, while also housing the necessary equipment for the fluorimeter under the themal PCR machine, as seen in our 3D model above. The fluorimeter system will have a handle that can pull up the lid to the fluorimeter box, keeping the sample in the dark for perfect fluorimeter conditions.

Current PCR process require two seperate machines to complete the PCR process, which makes the overall process tedious and sometimes faulty in nature. It requires the use of more materials which then leads to more clutter around the workspace, which may lead to some mistakes in the lab. Moreover, having two different machines to complete the lab will require more time since it will require the transferring of many samples back and forth between machines and camera adjusting.

Our machine solves these problems since the fluorimeter will have built in cameras inside the box so that the image of the droplets being tested can be easily converted and processed for easy information download. A major weakness in the PCR process was how long the overall process took and how much equipment we had to use, so to solve this problem, we combined the two aspects of the PCR process into one, easy to use machine. Moreover, the material that was used for the PCR machine in lab was wood, which could lead to problems such as malfunctioning equipment and breakage. However, our duoPCR solves this problem by utilizing a composite material that is not flammable.