BME100 f2015:Group9 1030amL6

From OpenWetWare

Jump to: navigation, search
BME 100 Fall 2015 Home
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
Wiki Editing Help



Name: Glenna Bea Embrador
Name: Glenna Bea Embrador
Name: Ian Quinn
Name: Ian Quinn
Name:Eric M Rhoades
Name:Eric M Rhoades
Name: Chase Gregor
Name: Chase Gregor
Name: Broderick T. Laese
Name: Broderick T. Laese
Name: Nimisha Tanna
Name: Nimisha Tanna


Bayesian Statistics

Overview of the Original Diagnosis System

The BME100 class tested patients through the disease-associated SNP by dividing the labor into 17 teams of ~6 students by having each team diagnose the total of 34 patients. By dividing the labor, determining the diagnoses of each patient became more accurate. In order to prevent error redoing the calculations was done as well as having each team do the calculation twice in order to double check everyone’s answers. In addition, since the images taken from the lab may not have been the best images, ImageJ’s calculations may have been off. In order to prevent this, several pictures were taken of the same DNA sample, ensuring that no outside source of light affected the sample. Another thing that was done in order to prevent any errors was performing each sample swiftly, yet carefully; making sure that micropipetting was accurate among other things in the lab. A challenge that we met was determining whether or not a patient had a disease SNP when comparing it to the ImageJ calculations, thus for the ambiguous values, some patients were considered inconclusive. For the class’s final data, some inconclusive results made the data slightly off, as well as the blank data, however from the data we were able to get an aggregate amount of affected and unaffected patients.

Calculation 1

Variable Description Numerical Value
APositive final test conclusion0.41
BPositive PCR reaction0.45
P(B|A)Probability of a Positive PCR reaction given that the patient receives a positive final test conclusion0.85
P(A|B)Probability of a positive final test conclusion given a positive PCR reaction0.77

Calculation 2

Variable Description Numerical Value
ANegative final test conclusion0.53
BNegative Diagnostic Signal0.49
P(B|A)Probability of a negative diagnostic signal given that the patient receives a negative final test conclusion0.82
P(A|B)Probability of a negative final test conclusion given that the patient receives a negative diagnostic signal0.89

Calculation 3

Variable Description Numerical Value
APatient will develop disease0.31
BPositive final test conclusion0.41
P(B|A)Probability that there will be a positive final text conclusion given that the patient develops the disease0.4
P(A|B)Probability that the patient develops the disease given that there is a positive final test conclusion0.3

Calculation 4

Variable Description Numerical Value
APatient will not develop the disease0.6875
BNegative final test conclusion0.53
P(B|A)Probability that there will be a negative final text conclusion given that the patient does not develop the disease0.54
P(A|B)Probability that the patient does not develop the disease given that there is a negative final test conclusion0.38

What Bayes Statistics Imply about This Diagnostic Approach

Within Calculation one, it describes the probability that given a positive PCR reaction, a patient will receive a positive final test conclusion. Since that probability is close to 75% as seen in the table above, this shows that there is a relatively high probability of this instance occurring, which makes sense because a positive PCR reaction will more or likely give a positive final test conclusion. But because this value is not as close to 100%, it shows that the PCR machine is not as accurate and a final “human” verdict would be needed in order to make a conclusion.

On the other hand, In calculation 2 it describes the probability that given a negative PCR reaction, a patient will receive a negative final test conclusion. Since that probability is close to 90%, this shows that the PCR machine is better manufactured at not detecting a disease SNP rather than vise versa. Due to the fact that this value is closer to 100% makes the machine more reliable, however as stated prior, it is not 100% accurate.

Regarding calculation 3, because the Bayes value is close to 30% and is less than 50% accurate, the ability of the PCR to diagnose and predict the development of the disease is not high. This result shows that although the PCR has a higher chance of detecting a disease SNP, it has a lower chance of diagnosis since other facts beyond it’s control can detect proper diagnosis.

In calculation 4, the Bayes value is close to 40% and similar to calculation 3, it is below 50% which further shows that the ability for the PCR to predict the development of the disease is not accurate.

Three sources of error include:

  • misdiagnosis/miscalculations by the teams, which qualifies as a human error since as a team we made conclusions on our respective patient.
  • The program ImageJ can only do so much with a given image, and if the image is not good quality, it can produce different numbers which may alter the data and thus altering the conclusion
  • Contamination can also affect data, and although measures were taken in order to prevent that, there may have been some contamination that may have affected the data samples.

Intro to Computer-Aided Design


Using TinkerCad was relatively simple after following the step by step tutorials that showcase all the tools possible. TinkerCad is user-friendly and despite being simple, had a variety of options in order to create different things. The first thing we did was follow the tutorial in order to understand the different tools such as zooming, the shapes, and grouping shapes together. After the tutorial, we downloaded Dr. Haynes files in order to begin redesigning the fluorimeter. Dr. Haynes files were much better suited for this particular lab compared to the basic tools given, however, when we were redesigning the fluorimeter so that the camera phone can take better images for data, the other shapes became useful. During the Computer-Aided Design we tried out different options and settled on creating a very mobile attachment to the fluorimeter so that the phone can get the best angle of the data. We used several shapes in order to complete the finished product. It was a bit difficult at first because if the shape is too big, it was hard viewing the whole device when the screen only showed so much.

Our Design


Our design keeps the basic box-shape of the fluorimeter, however, we have added an attachment (colorful portion) that makes taking down data much easier. The slider, which is in pastel pink, slides in and out of the fluorimeter in order to get as close as needed to the data that is being collected. The purpose of the sliding mechanism is to prevent the user from taking the phone out of the cradle constantly, and enable more mobile movement without disrupting anything. The light blue circles enable rotating ability of the attachment for more movement as well, and similar to a "selfie stick" the blue rod extends when needed in order to optimize the best viewing option. Finally, the light purple portion is where the phone is, and the difference from the OpenPCR design and this is that this design works like a clamp in order to secure the phone steady. On the other hand, in the OpenPCR lab, the cradle was unstable and the phone constantly fell or changed direction with slight movement which is why we opted for a design where we are able to position the phone and keep it at that position without it being moved.

Feature 1: Consumables

Very important consumables regarding PCR machines and any other biotech device is something that is an "additive" to the main device that can be thrown away and swapped for a new one. Since the fluoriotatemeter is the device where the samples are placed to record data the following things will be included in this device:

  • glass slides
  • plastic tubes
  • SYBER Green solution
Consumables​: plastics, pipettor, and reagents (PCR mix, primers) The OpenPCR​machine and software The Fluorimeter system​(including slides, stand, etc.)Numerical Value
STRENGTH: Cheap, durable, and already prepared respectivelySTRENGTH: Simple and easy to use. STRENGTH: Plenty of room to work with, besides the box preventing light exposure.
WEAKNESS: They could be easily misplaced without proper labeling.STRENGTH: Simple and easy to use. WEAKNESS: A more stable stand for phone/camera should be considered.

Feature 2: Hardware - PCR Machine & Fluorimeter

From Dee's Nuts n Bolts, The company that brought you the Tectonic Sole! The Fluriotatemeter!
Branding Positioning Target Markets Messaging Place
Customers who buy our product are buying a better functioning fluorimeter that enables the user to take better image data so more accurate conclusions can be madeWhat makes our product special is that it is more versatile with angles and accuracy in taking pictures that will be used as data for conclusionsOur customers consist of those who are searching for a more accurate and reliable fluorimeter that can help out with the difficulties of conducting an experiment with a regular fluorimeterAssuming our customers alreadyknow how to use a fluorimeter, our product requires no additional education at all! You just firmly place the camera on the clamp and rotate it to your desired photo angle The product can be bought online at, website where all labs get the best fluorimeters

Since propping and aligning the phone was difficult in the fluorimeter used in the PCR lab, the fluoriotatemeter fixes this by enabling more motility and better access to the phone to get data better.

Personal tools