One of the largest challenges that is faced by experts in the field of synthetic biology is the incomparable nature of fluorescence data as it has been reported in different units or processed in different ways. Our team participated in the Sixth International Interlaboratory Measurement Study and reproduced the scientific method as outlined by experiment one, identifying the likely sources of errors in fluorescence measurements and created a comparative analysis on Microsoft Excel between the three measured fluorescence from engineered bacterial constructs. For validity, all teams conducting this study have used E.coli K-12 DH5 alpha strain. Over the previous five InterLab Studies, the protocols investigated are considered comparable equivalents.
The protocols can be found here
Following the Experiment 1 protocol, fluorescence intensity of each construct was measured with 2 biological replicates, each of which had 4 technical replicates. Our team does not have access to the filters that completely match the wavelengths required by iGEM. Therefore, the readings were done with the available filters that are closest to requirements. Figure 3 shows the 3 filters and gain values used for fluorescence measurements. Gain values were adjusted accordingly to obtain the best signal possible. Calibration step was omitted due to inaccessibility to calibration beads.
For the blue and red fluorescence our raw data for the Interlab study plates at 0h and 6h timepoints. We found that in well number 10, the LB+C media contains erroneous data because it is an outlier out of the three other media and has similar OD to the blue fluorescence, which suggests that well 9 has contaminated well 10. However, the source of error for this data could also originate from the incorrect reporting or reading of the values. Thus, the average of well 1 and 12 was taken in the final results. The filter used for blue fluorescence was the incorrect filter, meaning that all the data for the blue fluorescence is invaild in this interlab.
The intensity of all fluorescents are determined by the number of cells or optical density (OD). As seen in Fig 2, the first and second colonies are replications of the same quantities and they have been measured four times, and averaged put in Fig 4, to reduce the impacts of errors.
At 0 hour time point for plate 1, where cell density is low for all test devices, fluorescence readings for all test devices were neglectable. After 6 hour incubation, the readings have varied significantly between different test devices. Figure 4 presents our data obtained from the three fluorescence measurements, where the yellow-highlighted rows are LB Cam, serving as blank.
Most of the constructs grew at a similar rate, except for the negative control and the two dual constructs, which are twice slower than the remainders.
The red and blue fluorescence readings have increased significantly; especially when compared to the green fluorescence. The difference in gain values may be a factor contributing to this discrepancy. However, when compared to the blank values of LB Cam, there were no test devices that had meaningful fluorescence intensity. This suggests a systematic error and calibration is crucial to obtain the correct results.
Interlab Study has been a good experience for us. Despite the equipment limitation as a high school team, and massive time constraints due to the last-minute arrival of the distribution kit, we gave our best attempt to complete the experiment. If time permits, more testing and repeats need to be done, in order to study the test devices accurately. The abnormality of the result also suggests the cruciality of the calibration steps, without which the collected data was not enough to draw many conclusions from.
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |