Model
This year, NWU-CHINA-A has designed a simple and portable detection device to detect the intestinal trace bleeding in non-experimental conditions. You can click there( https://2022.igem.wiki/nwu-china-a/hardware)to know more about this.
hardware
In order to gain further insights into whether the culture efficiency and stability of our hardware and the laboratory incubator differ greatly, we have cooperated with DKU_CHINA in modeling. We were responsible for providing the experimental data, and DKU_CHINA performed the data analysis. One experiment was conducted with the hardware as the experimental group along with traditional strain cultivation used as a control group. The data were then analyzed from the following two perspectives.
Time versus OD Regression
First, we tried to achieve a regression for time versus OD in order to gain further insights into the process of bacterial growth. After extensive trial and error, we obtained the following formulae for regression:
- Day 1 Control: y = − 0.0958x3 + 0.7311x2 − 0.9066x + 0.8140
- Day 1 Experimental: y = − 0.0662x3 + 0.5010x2 − 0.5510x + 0.6785
The corresponding graphs are as follows:
Control versus Experimental Regression
We also constructed a linear regression relationship between the control OD and the experimental OD. The result is :
- •Day 1: y = 0.7786x + 0.1486. R2 = 0.9294.
The plots are as follows:
The residual plots are as follows:
Based on those pieces of data we can conclude that the new hardware allows normal bacterial growth patterns despite that the strain concentrations are lower.
We concluded that:
- During calibration, the concentration of strains in the liquid used for calibrating was later found to be inconsistent, therefore indicating that the experimentally tested OD might deviate from the actual values.
- 2.Due to the problem that some of the team members in charge of operating the experiment lack experience, the operational errors in the experimental results are large.
Therefore, we reconducted an experiment to correct the shortcomings of the last experiment. The modeling results are as follows: