To begin the experiment, a total of 60 samples of DNA were taken from 20 patients. The samples were divided up evenly between 10 groups, giving each group 3 samples of two different patient's DNA. The three samples were transferred into test tubes, along with two more for a positive and negative control of DNA PCR mixture. The positive and negative controls are to eliminate the error. All of the tubes were then given the DNA primer mix to make a total of 100 microliters in each test tube. All of the samples were then ran through a PCR machine.
Prior to next part of the experiment, a fluorimeter was set up along with the iPhone to properly take the images for ImageJ. After that, 80 microliters of SYBR GREEN 1 were placed in between the first two holes in the slide from the fluorimeter kit, along with 80 microliters of the water blank solution (0 concentration of Calf Thymus DNA) and three pictures (that were made sure to be properly focused) were taken in the fluorimeter to compare the samples with a negative control. After that, the same procedure was done but with 5 samples of varying Calf Thymus DNA concentrations (ranging .25, .5, 1, 2, 5) to calibrate the ImageJ software and to reduce error. Next, the test tubes from the PCR machine were diluted into 8 separate test tubes with buffer in them. The same procedure that was done with the Calf Thymus DNA for calibration was performed for each sample of DNA that were in the PCR machine from the patients. Each sample from each patient had 3 pictures to be analyzed in ImageJ. After the data was calculated through ImageJ and Excel, the results of whether or not the sample had the disease or not were compared with each of the patients' actual diagnosis.
The class data shows that out of all of the samples, 19 tested positive for the disease from the PCR machine results. The results indicate that 39 of the samples tested negative, and 2 were inconclusive.
What Bayes Statistics Imply about This Diagnostic Approach
Calculation one, which determines the sensitivity for detecting the disease SNP, yielded a high probability of possessing the pathogen given a positive PCR result. This shows a good reliability in the PCR as a diagnostic system, but further tests are recommended for patients that had a positive result. This is because 1/4 of the positive patients still represent a significant number, and the calculations indicate that they may be misdiagnosed.
Calculation two, which on the other hand determines the specificity to detect the disease SNP, had a slightly higher percentage of patients not having the disease given a negative PCR result, compared with calculation one. This means that the system is more reliable to detect healthy patients rather than sick ones; further confirmation of the PCR results with other tests should be considered given these probabilities.
Calculation three determines the sensitivity to predict the disease, given the assumption of the correct diagnose by the doctor. The probability of the patient developing the disease SNP given a positive PCR result was the highest among all the calculations, with only 1/10 of the patients being misdiagnosed. This means the system is the most accurate for this type of result, predicting a development of the disease, which indicates a higher reliability and could be used as a final test.
On the other hand, calculation four, which determines the specificity for predicting the disease, had a lower probability. Specifically, this calculation, which determines the probability of the patient not developing the disease given a negative final conclusion, yielded incorrect diagnoses for 1/4 of patients. Even though this means that the system is accurate for most of the patients tested, it is not very precise. Further tests are recommended in the case of doubt given a negative final conclusion (as in calculation 1).
The main sources of human error for this experiment may have been: first, the incorrect amount transferred by the micropipette given the lack of expertise with the device; second, an incorrect capture of the photo, when using the fluorimeter, because of problems with the amount and align of light; third, contamination of the samples in the procedure that can lead to noise or alteration in the results.
Intro to Computer-Aided Design
3D Modeling
For the Computer-Aided Design Model, we utilized the TinkerCAD software. TinkerCAD was a nice alternative to Solidworks, as Solidworks is a more advanced and complex software. The software is more user-friendly and would be a great starting place to anyone who is new to creating 3D designs. Specifically, the software was extremely helpful for us when creating multiple shapes, resizing them, and moving them around to a precise position. Similar to GrabCad when we used Solidworks, Thingiverse for TinkerCAD was also useful when trying to create the drawer aspect of our renovation. Another aspect of TinkerCAD that is beneficial is that it is accessible online, so no additional programs need to be installed on one’s computer. Overall, we enjoyed the simplicity and ease of TinkerCAD’s software.
Our Design
Open PCR Machine:
Our design for the Open PCR machine did not change much from the original. However, to make the machine less bulky and inconvenient, we made it ⅓ smaller and plan to change the material of the machine to heat-resistant plastic. We also added a screen monitor, shown by the white rectangle, to take over the functions that the software on a computer would normally do, eliminating the need for excess equipment.
Fluorimeter:
Our new design shortens the time need to take a picture of each sample and analyze it. Firstly, instead of having a black box with a flap, we implemented a drawer containing the sample. The drawer is the rectangular black box frame and the black drawer itself in the design. The sample is the blue sphere on top of the rectangular slide. Once the sample is placed inside the drawer, the drawer would be closed. Inside the drawer would also be a built-in camera facing the sample, which captures pictures once the button on the outside of the drawer, the white half-sphere in the design, is pressed. The LED light would also be built-into the drawer. Additionally, a USB cable may be hooked up from a camera to an external device such as a laptop or phone to automatically have the pictures uploaded to the device to be analyzed.
Feature 1: Consumables
The cost of the consumables represented a setback in the design of the lab. Many of the materials are not reusable. However, for the micropipettor, it is best to use disposable materials to avoid cross contamination of substances. A good solution to minimize cost of consumables would be to manufacture a cheaper, less dense plastic with which the tubes and tips are made. The PCR mix and primers came in a box-like container that held small vials. Much of the space within the box was empty and remained unused, which represents a waste of materials since not all of the materials are being utilized. A solution to this would be to manufacture a smaller container in which the testing vials are placed and is made out of a cheaper, less dense kind of plastic. An improvement in the micropipettes such as having them with a certain fixed volume would prevent waste of the excess material. This adjustment would also reduce the chance of error in any of the measurements taken. For the glass slides being used for the drops, we plan to add labels on the circles to make it easier to keep track of the circles being used. Labeling the glass slides would help avoid mixing the drop with previous solutions or mixtures being used.
Feature 2: Hardware - PCR Machine & Fluorimeter
PCR Machine:
One of the strengths of the current PCR machine design is that it is simple and easy to use. However, the major downside to the current design of the PCR machine is that it takes a long time amplify segments of DNA via the polymerase chain reaction. To improve the design of the PCR machine, instead of using a software to set the parameters and configuration, a screen monitor should be incorporated into the machine so that the reaction could be performed directly. With this, there would be no need for a computer. Another vital change to ensure a more efficient design of the device would be to use a different material on outside of the PCR machine. Ideally, this material would be a type of heat-resistant plastic. This would make the machine lighter and easier to transport. The biggest issue with the PCR machine is the amount of time it takes to amplify segments of DNA. It is vital to test new ways to reduce the time of the process and overall reaction in order to ensure maximum efficiency. This could be done by using higher temperatures during the reaction or incorporating different reagents into the system. In order to further reduce the size of the machine, the inner components of the PCR machine, such as the heater and fan, for instance, would have to be compressed in order to reduce the size of the machine. Configuring the machine with standard temperature measurements for simplicity should also reduce the machine’s cost, size, and complexity.
Fluorimeter:
The current fluorimeter design is very simple, easy to use, and uses very basic technology. Although it contains very simple elements, it actually takes a very long time to complete the process. Due to this, the new design cuts down time and is a more stable environment for the sample. The design of our fluorimeter is also a black box, but with a very important twist. This box has a small, horizontal drawer that goes the length of the box. This drawer slides out to reveal the light stand where the sample is placed. The box also has a built in camera inside to capture photos. The change from the old design to this new one decreases process time by cutting the steps of lifting up the box housing the sample, finding a stand that properly hold the sample and is equal to the height of the smartphone used, and decreases risk of bumping sample when removing and placing the box over the sample. Yet another feature of the new fluorimeter is a built in camera. This aspect also decreases the process time because the user will not have to set a timer, reposition the smartphone each time a new sample needs to be recorded, or have to physically upload the images taken to a computer or device used to analyze the samples. Instead, they can be automatically uploaded to a technology device of the user's choice to eliminate time used to analyze photos of sample.
BME 100 Spring 2017
