Team:GDSYZX/Model
Experiment abstract
The pattern plant Arabidopsis and E. coli DH5α were used in this experiment . we use Arabidopsis protoplast to express our target genes and use HPLC y（high performance liquid chromatograph） to detect the targeted product.
In the first place, the genes 4HPAAS and UGT33 were optimized and synthesized according to the preference codon of Arabidopsis, We construct the expression vectors of 4HPAAS and UGT33: firstly, we conducted the enzyme digestion of vector pUC19-HA by BamHI and StuI. Then the two genes were cloned and recoveryed by DNA gel afterward. Secondly, we conducted Gibson Assembly to combine the target segments and the vectors. After that, the vector was transformed into E. coli DH5α and were selected by running PCR and sequence verification. After the expression vectors containing the target genes 4HPAAS and UGT33 were obtained, we extracted plenty plasmid and transformed them into protoplast of Arabidopsis . We first conducted western blot experiment to test the expression of protein and found both of them could expressed in protoplast of Arabidopsis.As the expression level of protein is corelated with the production of salidroside ,we then conducted western blot experiment to test the expression of protein at different times to acquire the best expression time of protein. After that, the target product was detected by HPLC. The results showed salidroside expressed in our protoplast. The result proves that our experimental scheme is feasible and we can reasonably to imagine a future in which the wide-used component presented in Chinese medicine, salidroside, is much less prohibitive than it is today.
Modeling
By Using Image J software to analyze the relative brightness of protein bands, we obtained the data of gray value of two genes/4h varying with time. There is a positive correlation between relative gray value and protein expression, so we use the standardized relative gray value to reflect relative protein expression level. Then we use SPSS software to draw the scatter plot of relative gray value about time to judge the overall trend. (The relative gray value is equal to the following gray value divided by the gray value of 4h)
We found that these scatters showed a quadratic function parabola trend, so we square the time data. Next, we test Pearson correlation coefficient of time, time squared and relative gray value to check whether there is a linear correlation between them.
Then we found that the correlation between UGT33 gene variables was quite high, P value was less than 0.05, with a high significance. The correlation coefficient of 4HPAAS gene was higher than 0.6, P value was higher than 0.05, but less than 0.2. Although the degree was not significant, the correlation coefficient was higher. It can be concluded that the binary linear regression equation can be constructed between the relative gray value of two gene sequences and the square term of time .
By using the least square method, we get the regression equation between relative gray value and time.
Then we analyze the fitting degree of the two models and calculate the R-square of the regression equation. We find that the goodness of fit of the two models is good, and the R-square is close to 1. Therefore, both models are better. In the case of significance level 0.05, the F-test of the two models is significant, which shows that all independent variable time has a higher linear significance for relative gray value on the whole, and the p-value of coefficient T test is less than 0.05. It also shows that each independent variable time has a higher linear significance for relative gray value.
From the fitting results, the relative gray value of UGT33 and 4HPAAS is proportional to time and inversely proportional to the square of time. From the results calculated from the model, it can be concluded that when Time=12.32, the relative gray value of UGT33 reaches a maximum value of 4.55; when Time=10.68, the relative gray value of 4HPAAS reaches a maximum value of 3.92. Since we measure every four hours, we can assume that when the time reaches about 9-12 hours, the expression reaches a maximum. It can also be seen from the two fitted images that the relative gray value increases with time and then increases and then decreases the parabolic trend.
Based on this modeling analysis, we conclude that 12 hours is sufficient for the experiment.
