BME100 s2016:Group16 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: Mayar AllamRole(s)
Name: Mayar Allam
Role(s)
Name: Tashena Jackson Role(s)
Name: Tashena Jackson
Role(s)
Name: Ryanne Maxie Role(s)
Name: Ryanne Maxie
Role(s)
Name: Andrew SmithRole(s)
Name: Andrew Smith
Role(s)
Name: Joseph CarrilloRole(s)
Name: Joseph Carrillo
Role(s)
Name: Austin MorganRole(s)
Name: Austin Morgan
Role(s)


LAB 6 WRITE-UP

Bayesian Statistics

Overview of the Original Diagnosis System


Overview of the Original Diagnosis System

In this experiment, 16 teams in BME 100 examined the DNA samples of 32 patients, and three

replicates DNA samples, summing up to 96 samples in total, were taken for each patient to

eliminate the sources of error, and have a good prediction of whether this patient will develop the

disease or not. The disease is mainly caused by single nucleotide polymorphism (SNP), so the

teams used PCR reaction to detect the SNP.

50 µL of each of the patients’ samples was added to a new tube, with 50 µL of the primer mix

plus the PCR mix added to each one. The tubes were then labeled properly, closed tightly, and

then placed in the OpenPCR Machine which has made the samples go through the repeating

cycles of the heating, and cooling cycles of PCR reaction to copy specific regions in the DNA.

The cycles start with the heated lid which was 100 Co, then the initial step which is at

temperature of 95 Co for 2 minutes. At this temperature, the DNA strands get denatures meaning

that each double stand DNA get separated in two single stranded DNA. Afterwards, the

temperature is lowered to 57 Co for 30 seconds, so the primers bind to complementary sequences

of DNA for, and at slightly higher temperature, which is 72 Co for 30 seconds, the enzyme TAC

polymerase binds to the primer sequences, and adds nucleotides to extend the strands, then the

samples are hold at temperature of 4Co, then the cycles get repeated over, and over to further

create more copies of the desired DNA. Positive, and negative controls were made as well in

order to compare all the DNA samples with them to see if the collected results are logical, and to

test if PCR has been done properly. Then, SYBER green were added to the samples, and the

samples were taken look at them in the fluorimeter box which emitted a light through the

samples’ droplets. 3 images were taken for each droplet per unique PCR sample to reduce

sources of error, and to get the most possible accurate results. Then, a calibration curve was set

up using the information collected from ImageJ for samples with known DNA concentrations.

Then, the calibration curve allows to know the DNA concentrations of the patients whose DNA

concentrations were unknown.

Among all the teams, only 2 samples were inconclusive, while there were other 30 successful

results. There were 12 patients with positive conclusion for the DNA test, and 18 other patients

with negative conclusion for the DNA test.

There were various sources of error in this experiment that may have affected the final results.

One of the major sources of error is the micro pipetting techniques. Since this experiment was

really sensitive to small changes, adding the improper amounts of the reactants might have

caused a great error shift.

What Bayes Statistics Imply about This Diagnostic Approach

Bayes Statistics

In calculation 1, the calculated probability that the patient will get a positive final test conclusion

was calculated, and found very small. Similarly, the calculated probability that the patient will

get a positive PCR reaction was found to be small as well. By using those probabilities, the

Probability that the patient will get a positive final PCR reaction given a positive final test

conclusion, and the probability that the patient will get a positive final test conclusion given a

positive PCR reaction was found to be very large; the values of those probabilities were found to

be near to 80%. Image:C11.png

In calculation 2, the probability that the patient will get a negative final test conclusion was

found to be near 50%, the same was for the probability that the patient will get a negative PCR

reaction which was found to be near to 50% as well. Given those probabilities, both the

probability that the patient will get a negative final PCR reaction given a negative final test

conclusion, and the probability that the patient will get a negative final test conclusion given a

negative PCR reaction were found to be very large, near to 90%, and 80% respectively.

In sum, it can be concluded that negative PCR result is more reliable in predicting a positive

disease conclusion than the positive PCR result due to the noticeable correlation between the

negative PCR result, and the final test conclusion. However, both the positive, and the negative

PCR results can be considered very reliable to predict the positive, and the negative final test

conclusions respectively. Image:C22.png


For calculation 3, the calculated probability that the patient will develop the disease was found to

be very small, less than 30%. Further, the calculated probability that the patient will get a

positive final test conclusion was found to be slightly higher, but still about to 40%. Using those

probabilities, the probability that the patient will get a positive final test conclusion given that the

patient will develop the disease was found to be exactly 50%, while the probability that the

patient will develop the disease given a positive final test conclusion was found to be less than

50%. It can be included that using PCR final result in predicting the development of the disease

are very unreliable. Image:C3.png

For calculation 4, the calculated probability that the patient will not develop the disease was

found to be very high, close to 70%, while the probability that the patient will get a negative final

test conclusion was found to be slightly less than that, which was close to 50%. Given those

probabilities, both the probability that the patient will get a negative final test conclusion given

that the patient will not develop the disease, and the probability that the patient will not develop

the disease given a negative final test conclusion were really high, close to 1.00.

In conclusion, the positive PCR results are not reliable in predicting the development of the

disease. However, the negative PCR results were very reliable in predicting the development of

the disease with much greater accuracy. Image:C4.png


In this lab there was a lot of room for error to enter the experiment, mostly in the image capture part of this experiment. The placement of the camera in a rather loose fitting stand caused constant movement of the camera, as well as the constant tapping of the phone to take photos. Light could have also easily entered the system as you only had three seconds to close the box completely and block out all light. The placement of the drop, in addition to light accidentally effecting the SYBR Green solution could also have changed the pictures and therefore the calculations.

Intro to Computer-Aided Design

TinkerCAD
Our group was for the most part proficient in SolidWorks, but the members who were not found the TinkerCad software and tutorials very helpful as an intro to computer aided design. The program ran rather slow for us due to poor internet connection, so after conferring with the professors we decided to instead download the STL files and design in SolidWorks. While these were older files and needed to be unzipped before they could be utilized, we found that designing in the software was easier and more reliable due to past experience. We took the current design and edited the design slightly to incorporate new features, and found that having the availability of these STL files to build off of was very useful. Our Design

Image:Openpcr123.png Image:Screen_Shot_2016-04-18_at_11.13.50_AM.png
Our design incorporates the original design of the OpenPCR machine and fluorimeter, with a few design edits in order to make it more effective and compact. We took the initial design and added a pin interface that allows someone to simply click on the fluorimeter to the side of the machine. The fluorimeter was more heavily redesigned as that is where we found the most issues. The new fluorimeter has a built in, self focusing camera that is placed a fixed distance of 2 centimeters from the droplet position. This camera has been calibrated to take more focused, clear and contrasting images of each droplet by aligning exposure, ISO, white balance, shutter speed and f-stop to take the best possible images. Built into the fluorimeter itself is a Blutooth component that allows the user to live view the droplet as it is under the blackbox, in addition to allowing them to take multiple pictures in succession. The fluorimeter also incorporates a slot that fits each slide in such a way that the top of the slide is perfectly flush with with surface of the fluorimeter, and there is an integrated blue laser light with a backdrop that shines the light directly through the middle of the droplet. The whole unit is coated with a silicon spray that prevents the drying and sticking of split liquids, and makes it waterproof. This in all makes the fluorimeter and OpenPCR machine combo more efficient, integrated, user friendly and compact.

Feature 1: Consumables

Crucial Consumables:

Standard glass slides

Primer solution

SYBR Green Solution

PCR mix

Buffer solution

Custom opaque black plastic tubes for SYBR Green solution

Standard Plastic tubes

Feature 2: Hardware - PCR Machine & Fluorimeter

The OpenPCR machine can be purchased or built with online schematics, and slightly altered in order to incorporate our new fluorimeter. The major redesign was the fluorimeter itself, especially the camera component, as most of the issues we encountered were from placing the camera. With this new fluorimeter, the camera is built in, along with Blutooth capacity that allows the images to be taken and recorded remotely. The user simply places the droplet on the slide, aligned with the laser, and covers the fluorimeter. The image live streams to an app on the phone, and snapshots can be taken remotely. This process eliminates the constant movement of the camera and blackbox, and allows for more consistent and accurate pictures. The entire unit is waterproof, and coated in a silicon spray that eliminates the need for major cleaning. We also redesigned the plastic tubes for the SYBR Green solution to be opaque, as to not affect the light sensitive solution.






Personal tools