Difference between revisions of "Team:CUNY Kingsborough/Results"
| (17 intermediate revisions by 2 users not shown) | |||
| Line 6: | Line 6: | ||
<h1 class="title-padding">Results</h1> | <h1 class="title-padding">Results</h1> | ||
| − | < | + | <h2 class="low-rise-padding">Introduction</h2> |
| − | <p class="low-rise-padding">To better characterize the standard curve for the EtBr-DNA Dot Blot, we are interested first whether the amount of EtBr significantly affects the fluorescent when the DNA concentration is varied between different amounts of EtBr and demonstrate there is significant difference in fluorescent between. We almost must characterize the predictive ability of the standard curve using some known DNA concentration that is not used for the curve.</p> | + | <p class="low-rise-padding">To better characterize the standard curve for the EtBr-DNA Dot Blot, we are interested first whether the amount of EtBr significantly affects the fluorescent when the DNA concentration is varied between different amounts of EtBr and demonstrate there is significant difference in fluorescent between. We almost must characterize the predictive ability of the standard curve using some known DNA concentration that is not used for the curve. Kasap et al. (2006) has shown that it is possible to approximate the concentration of DNA using Imagej. In this study we will use a linear regression model to find:</p> |
| − | < | + | <div class="no-rise-padding"> |
| − | <p class="low-rise-padding">Statistical analysis was done with R-studio and the Standard Curve was generated using Excel. Data was entered into Excel and standardized using the following fomula: | + | <ul> |
| + | <li>Is there a measurable pixel difference?</li> | ||
| + | <li>What are the factors that affect pixel intensity?</li> | ||
| + | </ul> | ||
| + | </div> | ||
| + | |||
| + | <h2 class="default-padding">Methods</h2> | ||
| + | <p class="low-rise-padding">Statistical analysis was done with R-studio and the Standard Curve was generated using Excel. Mass action model was done using Wolfram Mathematica. Data was entered into Excel and standardized using the following fomula: | ||
StandardPixel = (Value-MIN)\div(MAX-MIN)</p> | StandardPixel = (Value-MIN)\div(MAX-MIN)</p> | ||
| − | < | + | <h2 class="default-padding">Results</h2> |
| − | + | ||
| − | <h3 class="default-padding" | + | <h3 class="default-padding">Predicted Standard Curves</u></h2> |
| − | + | <center><img src="https://static.igem.org/mediawiki/2018/7/75/T--CUNY_Kingsborough--iGEMLinearStandardCurve.jpg" width="50%"></center> | |
| − | <p class=" | + | <p class="low-rise-padding">We examined the correlation of 125ng DNA and 16ng DNA under varied conditions of EtBr. We found that there is a low correlation when between the amount of EtBr and fluorescent.</p> |
| − | <h3 class="default-padding" style:"color: black"> | + | <h3 class="default-padding" style:"color: black">125 ng of DNA vs EtBr</h3> |
| − | + | ||
| − | <center>< | + | <center> |
| + | <figure> | ||
| + | <img src="https://static.igem.org/mediawiki/2018/a/a8/T--CUNY_Kingsborough--iGEM125ngDNAvsEtBr.jpg" width="50%"> | ||
| + | <figcaption><small>Fig 1. Scatter plot of 125ng DNA with varied amount of EtBr. r=0.23</small></figcaption> | ||
| + | </figure> | ||
| + | </center> | ||
| − | < | + | <h3 class="default-padding" style:"color: black">16 ng of DNA vs EtBr</h3> |
| − | < | + | <center> |
| + | <figure> | ||
| + | <img src="https://static.igem.org/mediawiki/2018/c/c7/T--CUNY_Kingsborough--iGEM16ngDNAvsEtBr.jpg" width="50%"> | ||
| + | <figcaption><small>Fig 2. Scatter plot of 16ng DNA with varied amount of EtBr. r=0.20</small></figcaption> | ||
| + | </figure> | ||
| + | </center> | ||
| − | + | <p class="low-rise-padding">When number of DNA was varied, we found that 125ng of DNA (β = 43.0, p < .001), and 16 ng of DNA (β=41.9, p < .001) were significant predictors of pixel intensity. Zero ng of DNA was used as the reference variable. The overall model fit was R^2 = 0.57, F-statistics = 41.02, p<.001.</p> | |
| − | + | <p class="low-rise-padding">This goes in contrast to our Mass-Action model which shows that the amount of EtBr has a significant effect on the pixel intensity. However, it is obvious that EtBr is needed for DNA to fluorescent therefore a future study should examine at what concentration of EtBr do we see no fluorescent when the amount of DNA is varied.</p> | |
| − | + | ||
| − | + | ||
| − | <p class=" | + | |
| − | + | ||
| − | + | ||
| − | + | ||
| − | + | ||
| − | + | ||
| + | <center> | ||
| + | <figure> | ||
| + | <img src="https://static.igem.org/mediawiki/2018/6/6a/T--CUNY_Kingsborough--resultsfig1.png" width="50%"> | ||
| + | <figcaption><small>Fig 3. Linear model results of lab data</small></figcaption> | ||
| + | </figure> | ||
| + | </center> | ||
| + | <p class="no-rise-padding">Figure 3 shows the linear model results for our own data. We found that amount of DNA (β=20.99, p<0.05) and EtBr (β=-0.38, p<0.05) are significant factors to pixel intensity. Background is not a significant contributor to pixel intensity. The overall model is significant (R2=0.387, F-statistics = 14.28, p<0.01).</p> | ||
| + | <center> | ||
| + | <figure> | ||
| + | <img src="https://static.igem.org/mediawiki/2018/9/98/T--CUNY_Kingsborough--resultsfig2.png" width="50%"> | ||
| + | <figcaption><small>Fig 4. Residual plot of figure 3 results</small></figcaption> | ||
| + | </figure> | ||
| + | </center> | ||
| + | <p class="no-rise-padding">Figure 4 shows the residual (Residual = Observed-Predicted). According to the residual, our model seems to under predict observed values and over predict lower values. This suggests that our model has not account for all the possible variables.</p> | ||
| + | <center> | ||
| + | <figure> | ||
| + | <img src="https://static.igem.org/mediawiki/2018/f/fa/T--CUNY_Kingsborough--resultsfig3.png" width="50%"> | ||
| + | <figcaption><small>Fig 5. Linear model from other teams</small></figcaption> | ||
| + | </figure> | ||
| + | </center> | ||
| + | <p class="no-rise-padding">Figure 5 is the linear model for the DNA-EtBr sample sent by other iGEM teams. DNA concentration (β=5.08, p<0.05) and background (β=-28.38, p<0.05) are significant contributors to pixel intensity. The results in Figure 3 suggests that the background-though we subtracted the background pixel intensity-is still a significant contributor to overall pixel intensity.</p> | ||
</body> | </body> | ||
| + | |||
| + | <br><br> | ||
</html> | </html> | ||
Latest revision as of 04:00, 8 December 2018
Results
Introduction
To better characterize the standard curve for the EtBr-DNA Dot Blot, we are interested first whether the amount of EtBr significantly affects the fluorescent when the DNA concentration is varied between different amounts of EtBr and demonstrate there is significant difference in fluorescent between. We almost must characterize the predictive ability of the standard curve using some known DNA concentration that is not used for the curve. Kasap et al. (2006) has shown that it is possible to approximate the concentration of DNA using Imagej. In this study we will use a linear regression model to find:
- Is there a measurable pixel difference?
- What are the factors that affect pixel intensity?
Methods
Statistical analysis was done with R-studio and the Standard Curve was generated using Excel. Mass action model was done using Wolfram Mathematica. Data was entered into Excel and standardized using the following fomula: StandardPixel = (Value-MIN)\div(MAX-MIN)
Results
Predicted Standard Curves
We examined the correlation of 125ng DNA and 16ng DNA under varied conditions of EtBr. We found that there is a low correlation when between the amount of EtBr and fluorescent.
125 ng of DNA vs EtBr
16 ng of DNA vs EtBr
When number of DNA was varied, we found that 125ng of DNA (β = 43.0, p < .001), and 16 ng of DNA (β=41.9, p < .001) were significant predictors of pixel intensity. Zero ng of DNA was used as the reference variable. The overall model fit was R^2 = 0.57, F-statistics = 41.02, p<.001.
This goes in contrast to our Mass-Action model which shows that the amount of EtBr has a significant effect on the pixel intensity. However, it is obvious that EtBr is needed for DNA to fluorescent therefore a future study should examine at what concentration of EtBr do we see no fluorescent when the amount of DNA is varied.
Figure 3 shows the linear model results for our own data. We found that amount of DNA (β=20.99, p<0.05) and EtBr (β=-0.38, p<0.05) are significant factors to pixel intensity. Background is not a significant contributor to pixel intensity. The overall model is significant (R2=0.387, F-statistics = 14.28, p<0.01).
Figure 4 shows the residual (Residual = Observed-Predicted). According to the residual, our model seems to under predict observed values and over predict lower values. This suggests that our model has not account for all the possible variables.
Figure 5 is the linear model for the DNA-EtBr sample sent by other iGEM teams. DNA concentration (β=5.08, p<0.05) and background (β=-28.38, p<0.05) are significant contributors to pixel intensity. The results in Figure 3 suggests that the background-though we subtracted the background pixel intensity-is still a significant contributor to overall pixel intensity.