BME100 f2016:Group10 W8AM 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
PRIDE Name: Eli Ozaki Name: Alexandra Dalbec Name: Nicholas Guzek Name: Mauro Udave PRIDE LAB 6 WRITE-UP Bayesian Statistics Overview of the Original Diagnosis System During BME100, students were split up into 16 groups that each tested two patients to see if they were positive or negative for a disease. Each group did three tests for each patient to see if the patient was positive or negative for the disease. The groups did each patient three times and took three pictures of each of the three results for both patients to minimize errors. All of the groups were instructed to use the exact same amount of drops for each patient to make sure everyone did the same process so results would be accurate. Every group also set up each of their calibration controls exactly the same. From the classes final results, only 14 of the 16 groups of data were on the spreadsheet. Throughout the spreadsheet, there were six results that came up as inconclusive. There were a total of 13 positive and 15 negative conclusions based on the DNA tests. According to our diagnosis data, if our test showed up to say that the patient had the disease, it was only about 70% accurate, whereas if it came up negative, it was about 75% accurate. So even though we all did the same procedure, the data was still not very accurate based on our results. There were a few things that could have possibly affected the data, such as groups possibly using to big or small of drops during the tests on accident, two groups data weren't added to the spreadsheet giving a smaller sample than if four more patients were added in, and groups could have misunderstood the directions in the workbook and done steps slightly different from other groups causing minor differences early on and end up getting into completely different results at the end. What Bayes Statistics Imply about This Diagnostic Approach Essentially, four calculations were made regarding the sensitivity and specificity of detecting and predicting the SNP with the original PCR machine. The first calculation found the probability of receiving a positive conclusion (being told the patient has the disease) assuming a positive PCR reaction, which is the sensitivity to detect the SNP disease, and the result was quite high at almost 90%. The second calculation found the probability of receiving a negative conclusion assuming a negative PCR reaction, which is the specificity to detect the SNP disease, and this result was also quite high at over 90%. The third calculation found the probability of actually getting the SNP disease assuming a positive diagnosis, which is the sensitivity to predict the disease, and the result was only moderately reliable at around 70%. The fourth calculation found the probability of not getting the SNP disease assuming a negative diagnosis, which is the specificity to predict the disease, and this result was also only moderately reliable at a little more than 70%. Obviously, this system does not appear to be as reliable as it should be. This can be a result of multiple sources of error throughout the process. For example, the SYBR green loses effectivity upon light exposure, so this component's reliability could have been influenced by being exposed to light on accident. Or, students running the test may have mixed the DNA solutions improperly, or even the PCR machine could have been calibrated incorrectly. Any of these and more possible errors could have influenced our Bayesian statistics findings. Intro to Computer-Aided Design 3D Modeling Our Design Feature 1: Consumables For the most part my group left the consumables the same because our focus was towards the PCR machine along with the Fluorimeter machine. The few items we did alter included the test tubes and the packaging of the tubes/tips, micropipettor, PCR mix, and primer. Were altering the packing of the materials so the materials come together instead of separately. Before everything is package we want the different tubes to be pre-labeled so the customer clearly understands which tubes are which, for example the customer can easily find the PCR mix because it will be labeled. Labeling the different tubes of reagents helps prevent confusion and can prevent incorrect data. The weaknesses my group identified from the consumables were the confusion of which tubes contained with reagents along with the fact that the primer had to be refrigerated unlike the rest of the materials, however my group couldn't seem to fix or improve that weakness so we left it the same unfortunately. Feature 2: Hardware - PCR Machine & Fluorimeter My group decided to alter the PCR machine so it seemed more convenient for the customer. First we decided that the machine was bulky and quite heavy so we reduced the size of the machine to the best of our ability and choose to use more of a lighter material for the outside panels of the PCR machine, we agreed on aluminum as our desired material. Next, we included more slots in the machine so more tubes can be placed in the PCR machine while making the results reliable and accurate. We left the program the same because we saw that the program was a strength because it was able to be connected to a computer and could be altered on there. As for the fluorimeter we designed the machine to have specific trays for each phone and make sure the trays are positioned at the right angle and distance for accurate measures every trial. The weakness we identified for the PCR machine was how the bulkiness of it, it was huge and heavy and not convenient for the owner. The PCR machine also limited the number of tubes allowed to be placed inside at a time. Instead of improving one aspect of the PCR machine we were determined to improve the quality and quantity of the machine, to do that we just included more slots allowing the quantity of the tubes to be placed in the machine to increase and improve the quality of the machine by using lighter materials and reducing the size of it. For the fluorimeter the weakness we identified was the process of positioning the tray and programing the phone being used, both of these factors played a key role in getting an accurate picture so we were determined to improve those factors of the fluorimeter machine.