BME100 f2017:Group7 W1030 L6
BME 100 Fall 2017 | 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 COMPANYPicture Perfect PCR LAB 6 WRITE-UPBayesian StatisticsOverview of the Original Diagnosis System Sixteen teams of approximately 5 students each diagnosed 2 patients total. To prevent error, there were 3 replicates per patient, a negative PCR control, a positive PCR control, and 3 images taken and analyzed in ImageJ that were blank control drops. Additionally, three drop images were used for each ImageJ calculation per unique PCR sample. In total, there were only two inconclusive conclusions, 17 positive conclusions, and 13 negative conclusions out of a total 32 DNA test conclusions. There was no blank data, though within the 96 total PCR reactions there was one inconclusive result. In total, there were 56 positive PCR results and 39 negative results out of 96. The frequency of positive PCR tests with positive conclusions within total positive PCRs was 49/56 or 87.5%. The frequency of negative PCR tests with negative conclusions within the total negative PCRs was 36/39 or 92.3%. For the diagnosis data, there were 11 positive conclusions with 'yes' diagnoses out of 17 total positive conclusions (64.7%) and 10 negative conclusions with 'no' diagnoses out of 13 negative conclusions (76.9%). Challenges the class researchers encountered that may have affected the data include using too little or too much SYBR green during the fluorimeter procedures and accidentally using a contaminated pipette with samples. This could have led to incorrect or inaccurate test results. What Bayes Statistics Imply about This Diagnostic Approach The results for calculation of sensitivity of the system with regard to the ability to detect the disease SNP was about 60%. The results for calculation of the sensitivity of the system with regard to the ability to predict the disease was close to 100%. These results indicate that though there was room for error with the sensitivity of detecting disease SNP, the test is highly reliable and sensitive to concluding that a person has the disease SNP. The results for calculation of specificity of the system regarding the ability to detect the disease SNP was about 80%. The results for calculation of specificity of the system regarding the ability to predict the disease was over 100%. These results indicate that the test was highly reliable and specific in regard to diagnosis. Three possible sources of error that could have occurred during the PCR and detection steps that could have affected the Bayes values are (1) Not fully pipetting the total sample amount onto the fluorimeter slide and thereby creating an incorrect concentration ratio, (2) mixing up the droplet fluorimeter images related to specific patients and replicates, and (3) not having a constant distance between the phone camera and the droplet across all replicates. In all of these cases, incorrect values could have led to incorrect calculations that might have given a positive, negative, or inconclusive result when that wasn't actually the case. This could have made it seem like the sensitivity and specificity were either more or less reliable depending on how the error affected values and calculations. Intro to Computer-Aided Design3D Modeling The software our team used was Tinkercad, as we felt the online interface would be easier to manage from a team perspective than SolidWorks. Our experience on the whole was quite positive. We found the import system easy to use and the commands simple to understand. We started out by getting familiar with the viewing, selection, and manipulation tools. The group and align tools were especially useful for creating holes and combining shapes into one final merged design. For the PCR machine, we utilized the parts provided by Professor Haynes for constructing the 3D model. Then, we re-designed the fluorimeter using Tinkercad basic shapes adjusted to specific sizing parameters. Another function we utilized heavily was the workplane, which enabled us to add parts to the sides of other parts in 3D space. Additionally, Mac keyboard commands like 'ctrl with right click' to rotate an object were very useful throughout the design process. Our Design
Feature 1: ConsumablesIncluded in our consumables kit will be:
The listed materials are required in order for the lab to maintain accurate and reliable data that can be collected using our machines. The PCR reaction mix is pre-made and can be used in conjunction with DNA/primer mix provided by the lab. The 8 tubes are designed specifically to fit the heating block for our PCR machine. Also, the hydrophobic slides would be included in the kit because they are designed to fit our specially designed fluorimeter. Additional supplies for running a PCR reaction such as pipettes and pipette tips would likely already be owned by the lab. Our consumables kit accounts for the supplies that a lab may not have on hand for the entire experimental procedure. Feature 2: Hardware - PCR Machine & FluorimeterThe Open PCR machine and the fluorimeter will both be included in the system. They will be packaged in their complete assembled form separately alongside the consumables outlined above. Our fluorimeter design does not require any separate pieces and everything is attached to one central unit (no separate phone cradle or black box). The Open PCR machine and the fluorimeter will both be single units and are not connected to one another. Our hardware will include:
Our group decided to redesign the fluorimeter because with the original design it was difficult to get consistent and accurate pictures of the droplets. The cradle was easy to bump and move out of position. Additionally, there was no fixed height at which the fluorimeter sits in comparison to the size of the phone. Thus, the machine was not built for a specific phone brand, which makes sense for a small-scale student lab but would lead to inconsistency and inaccuracy in a full-size lab. We re-designed the fluorimeter to make taking pictures occur at an exact height and distance from the droplet. Our design is made custom for a Samsung S8's standard dimensions (148.9 mm × 68.1 mm × 8.0 mm). The cradle positions the phone 5 centimeters from the edge of the slide holder. The main fluorimeter body is positioned at the ideal height for the phone's camera. If we were to actually manufacture the machine, we would likely create a few different designs for different common phone brands. The accuracy and consistency of the droplet pictures taken with our device would be excellent because the cradle is always the same distance from the droplet. Our fluorimeter also has an adjustable flap so that a separate black box covering is not needed. |