BME100 f2016:Group1 W1030AM L6

BME 100 Fall 2016	Home People 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 Photos Wiki Editing Help
OUR TEAM Carlye A. Frisch Kendra M. Gibble Anna Rothweil Gabrielle F. Wipper Nicholas C. Whitley The PCR Popper LAB 6 WRITE-UP Bayesian Statistics Overview of the Original Diagnosis System In the BME 100 class, 16 teams of 4-5 students each, diagnosed a total of 32 patients. Each team was given 2 patients to diagnose. In our group, one team of 3 did the PCR reaction and another team of 2 focused on the calculations. To prevent error, both a positive and a negative control were used for comparison of the solutions. For each patient, 3 samples of solution were used for more trials in order to prevent error caused by not taking enough samples. With the ImageJ calibration controls, an oval-shape was saved to the ROI manager for each sample so the 3 images for each sample would have the same oval measurements. Also, the green color split image was measured for each of the images. For each concentration of solution, 3 images were used to prevent error in the calculations. The class's final data concluded that there were more negative diagnoses than positive ones. The class had 3 inconclusive results and 4 blank data results were found for the class's final data as well. Group one's data may have been affected by an accidental contamination of the DNA which would cause the data to not portray the conclusions this group received. Probability of a POS final test conclusion, given positive PCR reaction Variable Description Numerical Value A Frequency of POS conclusions 0.269 B Frequency of pos individual PCR 0.295 P(B/A) Given POS conclusion, what is probability of individual pos result 0.87 P(A/B) Given pos individual result, what is probability of POS conclusion 0.793 Probability of a NEG final test conclusion, given negative PCR reaction Variable Description Numerical Value A Frequency of NEG conclusion 0.615 B Frequency of neg individual PCR 0.603 P(B/A) Given NEG conclusion, what is probability of individual neg result 0.936 P(A/B) Given neg individual result, what is probability of NEG conclusion 0.955 Probability of patient disease, given POS final test conclusion Variable Description Numerical Value A Frequency of Disease 0.269 B Frequency of POS conclusion 0.269 P(B/A) Given Disease, what is probability of POS conclusion 0.429 P(A/B) Given POS conclusion, what is probability of Disease 0.429 Probability of not patient disease, given NEG final test conclusion Variable Description Numerical Value A Frequency of Not Disease 0.731 B Frequency of NEG conclusion 0.615 P(B/A) Given Not Disease, what is probability of NEG conclusion 0.684 P(A/B) Given NEG conclusion, what is probability of Not Disease 0.813 Sensitivity is the ability of the test to identify those with the disease--a true positive. Specificity is the ability of the test to identify those without the disease--a true negative. The sensitivity of the system to detect the disease SNP is described by: Table 1 The sensitivity of the system to predict the disease is described by: Table 3 The specificity of the system to detect the disease SNP is described by: Table 2 The specificity of the system to predict the disease is described by: Table 4 What Bayes Statistics Imply about This Diagnostic Approach In the first two calculations, finding a negative individual result very likely led to a NEG conclusion, almost 100%, and vice versa. For the positive calculations, it was more like 80-90% that it would be a positive individual result and a POS conclusion, and vice versa. However, it was noted that there were far more misses by the class in individual tests, when the doctors had come to a POS conclusion. This could be explained by a few different errors and types of errors, which is discussed below. The predictions for developing the disease, or not, in calculations 3 and 4 were not as strong. Having the disease and predicting it, or vice versa, shown in table 3, had a very low probability, less than 50%. It seems this is at least in part due to the fact that the class found 7 positives, and there were 7 positives in all of the patients (out of 26 that were tested with conclusions), but, many of these were wrong. Sometimes positive was predicted and was wrong, and sometimes positive was not predicted, but should have been. The negative predictions were more accurate, around 70-80%, but again, this could largely be due to the number of negatives overall in the test sample. One aspect of human error that could have occurred during the PCR and detection steps include the amount of outside DNA that might have possibly ended up in the sample. If any outside DNA fell into the sample, values could have had possible error, affecting the Bayes values in a negative way. Another possible source of machine error was the amount of liquid that actually came out of the pipette. At some points during the experiment, there were some drops left in the pipette that could have overall affected the results of the experiment. The third possible source of error was the camera. The phone camera could have moved, or become unfocused at certain points of data recording. This would have affected the images in ROI manager, which in turn could affect the overall results of this experiment. Intro to Computer-Aided Design 3D Modeling Our team used Solidworks for this particular lab. We found Solidworks to be a difficult program to use if there is little to know previous experience. It took us a long time to understand the controls on this program. However, when we learned how to use Solidworks we found the designing process to be much easier. It allowed us to see the system and manipulate it in order to add our alterations. (Any picture can be clicked on to zoom in) SolidWorks PCR Machine Close-up of the new power supply Full view of the new power supply View of current PCR machine with power supply and mock up of new power supply Side view of current PCR machine and power supply After speaking with Professor Haynes, she told the group to move forward with the idea, showing pictures of the device as a comparison to the original power supply. Our design includes a much smaller power source. The goal the group settled on was to reduce the size of the overall unit, while still allowing for the same number of simultaneous tests. As you can see from all of the above pictures, the new power supply is considerably smaller than the current option. It plugs into the board in the exact same spot (where all the colorful wires are coming out of the white plastic piece plugged into the board) and will quite literally be attached to the board, sticking up about two inches. Looking closely, the yellow tips of the new power supply will go where the white plastic and cords are sticking out of the board. The other cords attached to the new power supply plug into various places, such as the current fan and heat sink. However, a large part of the hot air that is introduced into the system comes from the power supply itself. By removing that, there may be further improvements possible with different fan and heat sink setups. For now, the goal of the group is to replace the giant gray power supply with this new very small design. As one can see, it is compatible with the devices and power supply requirements currently in use in the PCR machine, and it has a small wire that allows a power cord, similar to nearly any laptop power cord on the market, to plug in on an external wall of the device. Feature 1: Consumables A major weakness found with the PCR packaging was its organization considering the amount of steps required to complete it. To combat this problem, in the optimized design, the original PCR lab consumables will remain the same. Included is the PCR mix, SYBR Green Solution, micropipettor, buffer, liquid reagents, glass slides, plastic tubes and the pipette tips. The consumables will be organized in the order they are used in the PCR lab. A pipette tip will be arranged first, followed by the PCR mix, another pipette tip and so on corresponding with the various steps necessary for the PCR lab. This new method of packaging will make the process of completing the lab much easier for each user. Feature 2: Hardware - PCR Machine & Fluorimeter Our PCR machine has nearly the same hardware as the original Open PCR, but has a much smaller power supply. The replacement of the power supply is the major innovation. As a consequence, not only the size of the Open PCR machine is decreased, but also its weight is significantly smaller. The name of the new power supply is The PCR Power Popper (Exclusively used in the PCR Popper design) A dark box with a fixed fluorimeter in it will be included with our device. It can be used to analyze the samples prepared with the Open PCR machine with the help of a smartphone. There is also innovation within, where there is an adjustable stage inside the box, about 10 cm away from the fluorimeter. The stage can be adjusted so that the phone can be raised or lowered, and set farther from or closer to the fluorimeter. As seen from the pictures and descriptions above, the biggest change is to the the PCR machine, now called the PCR Popper (from the people who brought you the Pill Popper) because of the addition of a redesigned, very small power supply, called the PCR Power Popper (used exclusively with the PCR Popper). This will make the machine much smaller and likely more efficient and durable because of the removal of much of the heat inside the unit (just as one would see with a computer that runs cooler). In addition, the phone stand for the flourimeter will be adjustable to make it easier to set up, and get the right viewing angle for the phone camera.