BME100 s2016:Group4 W1030AM L6

From OpenWetWare

Jump to: navigation, search
BME 100 Spring 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
Image:BME494_Asu_logo.png

Contents

OUR COMPANY

Name: Paul Gossett
Name: Paul Gossett
Name: Desiree Yazzie
Name: Desiree Yazzie
Name: Sara Gubrud
Name: Sara Gubrud
Name: Katerina Soltero
Name: Katerina Soltero
Name: Patrick Panattoni
Name: Patrick Panattoni
Name: Karson Pooler
Name: Karson Pooler


LAB 6 WRITE-UP

Bayesian Statistics

Overview of the Original Diagnosis System

The BME 100 class used a disease SNP diagnostic system to diagnose patients. The division of labor was shared between seventeen groups with six students. Overall 34 patients were diagnosed.

In the beginning of the lab, each group received three replicate DNA samples from two patients. The number of replicates per patient were done in order to prevent error, because three results are better than two. More results gives reliability to the accuracy of results.

PCR controls were done by capturing the fluorescence of DNA with a camera phone. Three images were captured for each PCR reaction so the data could be more reliable. An iPhone was used because the optics was state of the art. SYBR green I was used to help capture the fluorescence when used in a fluorimeter.

The ImageJ calibration controls helped determine the fluorescence of Calf Thymus by comparing the RAWINTDEN of the drop and the RAWINTDEN of the background. Statistics were used to find the mean and standard deviation of the RAWINTDEN of the drop minus RAWINTDEN of the background for three images. The data for the calf thymus helped generate a calibration curve. The calibration curve helped determine DNA concentration of each PCR reaction.

The class's final data was shared on a spreadsheet. There were few inconclusive results which was expected. Each group had "yes" or "no" results. If a patient had three "no's" then the patient did not have the disease, and vice verse. Then Bayesian statistics was used to determine the reliability of the diagnostic tool. Such as what is the percentage that a patient receives a negative result if he or she really does have a disease. Any challenges were using new methods of analyzing data.


What Bayes Statistics Imply about This Diagnostic Approach


Bayesian statistics can be used to determine the reliability of the PCR reaction test to detect disease SNPs and the accuracy of the test to predict whether or not a patient will develop the disease. Our Bayesian statistics results indicated that the probability of getting a positive final test conclusion given a positive PCR reaction was close to 1. The probability of getting a negative final test conclusion given a negative PCR reaction was close to 1. These high probabilities mean that the PCR reaction test can reliably detect disease SNPs.

The probability of the patient developing the disease given a positive test conclusion was under .5. This low probability indicates that the accuracy of the PCR reaction test was low. There were many false positives which means that many patients who received a positive test conclusion did not develop the disease. The probability of the patient not developing the disease given a negative test conclusion was over 1. The probability over 1 indicates that the PCR reaction test was not accurate at predicting whether or not a patient will develop the disease because probabilities should not be over 1. There was a lot of false negatives where the patient received a negative test conclusion but developed the disease. In conclusion, the PCR reaction test can reliably detect disease SNPs but is not accurate at predicting whether or not a patient will develop the disease.

One source of human error that could have occurred was a concentration calculation error when determining whether or not a PCR reaction test was positive or negative. This would result in inaccurate test results and therefore inaccurate Bayesian statistic results. Another source of human error could have come from students just eyeballing the test results as opposed to actually calculating the concentrations. This would also result in inaccurate Bayesian statistics results. Lastly, a source of machine error was that the camera was not accurately capturing the light flowing through the sample. This could result in inaccurate concentrations and therefore inaccurate test conclusions. These errors were minimized by taking three pictures per sample but some error could still remain.

Intro to Computer-Aided Design

TinkerCAD
Image:Screen_Shot_2016-04-13_at_11.28.03_AM.png‎

Tinker CAD was a useful little tool. However, it does not compare to the mobility that SolidWorks allows. SolidWorks allows for more views when assembling a multiple part device. It was very simple to import parts for the PCR machine which was very useful. Tinker CAD also allowed us to change each parts color however it did not use different material types that SolidWorks allows.

Our Design


Image:Screen_Shot_2016-04-13_at_11.45.29_AM.png‎Image:Screen_Shot_2016-04-13_at_11.44.59_AM.png


Our PCR machine contains a light resistant, cyber storage. It includes pre-labeled tubes to ensure samples do not get mixed up. The tubes are also easy to break apart for separate storage. Our PCR machine has more sample storage so that more samples can be ran at the same time. Our PCR machine also includes an adjustable cradle to ensure that ALL phone types and cameras fit. The first photo above, is the PCR machine our design was based off of.

Feature 1: Consumables

  • Pre-labelled tubes
  • Perforated tubes that can easily be broken apart
  • Light resistant storage for SYBR green


The features listed above addressed the weaknesses in the current diagnosis system. We struggled with labelling the tubes because they were so small and the permanent markers easily smeared. Our feature of pre-labelled tubes allow for easy tracking of the samples. we also found it difficult to break apart the tubes because you had to twist the tubes until they broke apart. Lastly, we struggled with keeping the SYBR green under the aluminum foil, especially when we were trying to work quickly. Our feature of a light resistant storage unit for SYBR green allows for the elimination of the aluminum foil. The users do not have to worry about light destroying the SYBR green and ruining the whole PCR reaction test.

Feature 2: Hardware - PCR Machine & Fluorimeter

The PCR reaction is a complex cycle of heating the samples at a specific temperature for a certain amount of time. The group can redesign the PCR machine that could time the reactions and set the desired temperatures. The redesigned PCR reaction can also store more reactions, and therefore give more results.

The fluorimeter was difficult because the iPhone we wanted to use couldn't stay in one place. So the redesigned fluorimeter would have a built in camera or either a camera dock that could be adjustable. This necessary so a person could save time by not tinkering with the fluorimeter. Also the camera will be able to stay in one place so the distance from the camera to the drop is constant.







Personal tools