After months of research, all my work came together into a poster that I presented at the Zebrafish Disease Models Society conference (ZDM8). At that conference, I listened to many exciting lectures on the cutting edge of Zebrafish science, including cancer research, stem cell research, and spinal regeneration. Afterwards, I presented my work to researchers from all across the world. After many questions and encouraging comments, I came out with a much better understanding of which direction I should be taking this machine.
Photo of me at the conference:
One of the most important steps when producing our thermocycler is producing a printed circuit board (PCB). A PCB allows for the mass production of our circuit board for little to no effort and for a low cost. After hours of work throughout the summer, we finally received our first PCB back from the production facility in Germany. Placement of the parts on the board and programming is currently underway.
Image of PCB from production facility:
Before the upcoming ZDM8 Conference in Boston, an extensive comparison between my thermocycler and conventional thermocyclers needed to be performed. When comparing, I focused on the key aspects of thermocyclers. These aspects are weight, cost, and energy consumption. After comparing my thermocycler with five other conventional thermocyclers in each of these aspects, the following graphs were produced:
After constructing the circuit for the detection system with the raw electrical components, I sought preliminary validation of the detection system to determine the feasibility of this design.
Preliminary Testing Setup:
I first set up one light sensor with two LEDs and an optical filter to test one detection unit. I then performed qPCR on 4 positive samples of DNA and 2 negative DNA samples. Afterwards, I performed two simple tests. First, I looked to see if the DNA could fluoresce. Secondly, I judged the specificity of the light sensor to determine if it is capable of detecting the small changes in the DNA's fluorescence.
Firstly, I determined that the DNA did indeed fluoresce when using the SYBR green fluorescent probe under a 480 nm LED. Secondly, I determined that the Cadmium Sulfide optical sensor could detect very small changes in the light intensity. This was accessed by observing a millivolt level of specificity when using this sensor in a voltage divider circuit under a variable light source. Lastly, I determined that I needed to purchase higher quality LEDs because the optical filters revealed that our current LEDs emit a range of wavelengths of light.
After choosing to develop a fluorescent detection system, I spent several days researching and understanding the biology of the detection system as well as the technical approach that conventional qPCR machines are taking to construct their detection systems. After a week of designing the mechanical and electrical layout of my new low cost, low power fluorescent detection system, the following design was produced:
This design utilizes a light sensor with an optical filter flanked by two LED's with wavelengths set to that of the intended fluorescent dye that needs to be fluoresced. After designing the detection system, I spent some time integrating this detection system into my design and produced this low cost, low power qPCR machine:
By combining the fluorescent detection system with my PCR machine, I effectively created a qPCR machine. This qPCR machine will be utilized as a low cost, low power genotyping tool for genetic research, genetic testing, and disease detection.
How it works:
After the extension phase of PCR each cycle, the test tubes are scanned through this detection system for the presence of fluorescence. By analyzing the changing levels of fluorescence throughout the PCR amplification, I can effectivity determine both the amount of a targeted molecular material that is present along with the simple binary presence of a disease or piece of molecular material.
Another important aspect I needed to focus on this summer when developing my thermocycler was the user interface. Over the summer, a preliminary interface was developed using Java. This interface allows for temperature programming, debugging, and testing.
Image of preliminary thermocycler interface:
After developing the 5th prototype of the thermocycler with confidence in its performance from the previous verification of the control algorithm, a final detection system still needed to be chosen. In order to select the final detection system, I utilized the Standford Launchpad program to help create a business plan for my product.
This final business plan used was the following:
This business plan allowed me to fully understand my product and how it could function with each plausible detection system. The detection systems that I considered were the following: electrophoresis, micro capillary electrophoresis, a fluorescent detection system, and a portable strip test. Ultimately, a fluorescent detection system best paired with my low cost, low power thermocycler. Development of the fluorescent detection system is underway.
After extensive work over the summer, a compact and reproducible prototype was developed to integrate the previously validated temperature control algorithm with 3D printing capability and a more compact design. A 3D model of this design can be found below:
This machine features two major differences that differentiate this design from the previous design:
Some minor changes are listed below:
After extensive testing of the fourth prototype, success was yielded. First of all, temperature validation was performed on my thermocycler using temperature sensors. This validation revealed that this prototype had a high degree of accuracy as well as temperature uniformity among the wells in the thermocycler. Secondly, a testing assay was performed to compare this thermocycler's accuracy with a conventional thermocycler's accuracy. Results showed a comparable level of accuracy between the thermocyclers, indicating that this thermocycler is capable of replacing laboratory grade thermocyclers without sacrificing a significant degree of accuracy. However, a more reproducible and compact version of the machine is desired, so the development of the 5th prototype is underway.
Figure 1 features the electrophoresis results from a conventional thermocycler.
Figure 2 features the electrophoresis results from my thermocycler.
Results: The bands produced from my thermocycler were comparable to those produced from the conventional thermocycler, although one column was missing due to loading errors preceding the electrophoresis test.
This figure features 3 lines that plot the temperatures of the three separate oil baths as they maintained their delegated temperature throughout PCR.
Results: Throughout the entire PCR test, the averaged temperature variation was +- .2 degrees Celsius throughout the wells in the thermocycler.
Currently I am developing my 4th Prototype of my Thermocycler. By improving the control systems, the temperature accuracy has significantly improved! I also have reduced the price of the machine and am working on making a 3-D Printable Version, as well as a customization version of the machine. Soon, I am going to test this machine in a laboratory.