BME103:T130 Group 15: Difference between revisions
Nehal Jolly (talk | contribs) |
|||
Line 152: | Line 152: | ||
This is all in theory of course, and should work perfectly every time. But there are many factors to consider. For instance, not every one who has cancer has the cancer gene. | This is all in theory of course, and should work perfectly every time. But there are many factors to consider. For instance, not every one who has cancer has the cancer gene. | ||
<br><br> | <br><br> | ||
'''Reliability and Accuracy of Specific Cancer Marker Detection'''<br> | |||
Based on Bayesian reasoning, the probability that someone will test positive that will actually have the disease has approximately 7.8 percent. However, The chance that someone does not test positive, and doesn't have cancer is about 99.8%. | |||
These four percentages generally mean that the test itself is generally in favor of testing negative, which means there are less chances to have false diagnosis and/or treatments. | |||
<br> <br> | <br> <br> | ||
These percentages resulted from the following equation:<br> | These percentages resulted from the following equation:<br> | ||
Line 168: | Line 164: | ||
The values for these above equations are gathered from Bayesian's reasoning. | The values for these above equations are gathered from Bayesian's reasoning. | ||
Baye's Theorem: <br> | Baye's Theorem: <br> | ||
<math>p(C/T) = (p(T/C)*p(C))/(p(T/C)*p(C)+p(T/nC)*p(nC))</math> <br> | |||
<math>p(C/T) = (p(T/C)*p(C))/(p(T/C)*p(C)+p(T/nC)*p(nC))</math> <br><br> | |||
where p(C/T) is the probability of a person with positive results will have cancer out of the entire patients participating. p(C) is the probability of having cancer present, p(T/C) is the percent of patients who tested positive with have cancer and had it. n = not | where p(C/T) is the probability of a person with positive results will have cancer out of the entire patients participating. p(C) is the probability of having cancer present, p(T/C) is the percent of patients who tested positive with have cancer and had it. n = not | ||
<br><br> | <br><br> | ||
The values of the these equations are taken from the general statistics from tests performed on 'x' amount of patients. For instance, the direct results showed that 80% of the people with cancer, tested positive, this value would be used as p(T/C) because that is the test running positive and having cancer. The calculated p(C/T) was used as the "True Positive" in the original four equations. The same for the false positive, true negative and false negative can be solved similarly as well. | The values of the these equations are taken from the general statistics from tests performed on 'x' amount of patients. For instance, the direct results showed that 80% of the people with cancer, tested positive, this value would be used as p(T/C) because that is the test running positive and having cancer. The value p(T/nC) would be the percentage of people who tested positive but do not actually have cancer, which resulted to be 9.6%. Additionally, the general statistics were 90.4% of patients will test negative and will not have cancer, and 20% percent of people with cancer will run negative. | ||
<br> The calculated p(C/T) was used as the "True Positive" in the original four equations. The same for the false positive, true negative and false negative can be solved similarly as well. | |||
<br><br> | <br><br> | ||
Revision as of 18:09, 14 November 2012
BME 103 Fall 2012 | Home People Lab Write-Up 1 Lab Write-Up 2 Lab Write-Up 3 Course Logistics For Instructors Photos Wiki Editing Help | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
OUR TEAMLAB 1 WRITE-UPInitial Machine TestingThe Original Design Experimenting With the Connections
ProtocolsPolymerase Chain Reaction
The Components of the GoTaq® Colorless Master Mix
DNA Samples (8)
(Add your work from Week 3, Part 2 here aka, steps of assembly to the flourimeter)
Research and DevelopmentSpecific Cancer Marker Detection - The Underlying Technology There is a genetic relation to having cancer or not when an individual is over the age of 40. The specific gene, in this case, is located on chromosome 22, r17879961. To test an human's DNA for this cancer gene, we have go through a series of reactions called PCR on the DNA for replication and amplification of the patient's DNA strand.
Based on Bayesian reasoning, the probability that someone will test positive that will actually have the disease has approximately 7.8 percent. However, The chance that someone does not test positive, and doesn't have cancer is about 99.8%.
These four percentages generally mean that the test itself is generally in favor of testing negative, which means there are less chances to have false diagnosis and/or treatments.
[math]\displaystyle{ Negative Predictive Value = (True Negative)/(True Negative + False Negative) }[/math] The values for these above equations are gathered from Bayesian's reasoning.
Baye's Theorem: [math]\displaystyle{ p(C/T) = (p(T/C)*p(C))/(p(T/C)*p(C)+p(T/nC)*p(nC)) }[/math] where p(C/T) is the probability of a person with positive results will have cancer out of the entire patients participating. p(C) is the probability of having cancer present, p(T/C) is the percent of patients who tested positive with have cancer and had it. n = not
Results
|